Key Issues In Methadone Maintenance Treatment. Chapter 4 "Treatment Effectiveness III: AIDS and Other Infectious Diseases"

i. Introduction

An important aspect of the effectiveness of methadone maintenance (MM) is the extent to which it prevents the spread of infectious diseases among injecting drug users. The massive expansion of methadone maintenance programs in Australia and other countries in the latter half of the 1980s has taken place as an attempt to prevent the rapid increases that have been seen in some major urban centres throughout the world (e.g. New York, Edinburgh, Milan, Bangkok) in rates of infection with the human immunodeficiency virus (HIV) among injecting drug users. The two variants of the human immunodeficiency virus are known as HIV-1 and HIV-2. The most common variant infecting drug users throughout the world is HIV-1, although HIV-2 infection, which is more common in African countries, has also been reported (for a full discussion see Grmek, 1990). In this book, we use the generic term HIV which, at the time of writing, means HIV-1.

Although they have received less attention, other infectious diseases such as hepatitis B and C are also important in any consideration of protecting the health of drug injectors. For an interesting account of why the hepatitis B epidemic received much less attention than HIV see Muraskin (1988). Our discussion will focus on HIV because there is a large and rapidly growing literature on this topic. While the acquired immune deficiency syndrome (AIDS) poses a public health problem without precedent in recent times, we emphasize that it is only one of a number of serious health consequences associated with illicit drug use (Selwyn, 1991; Selwyn et al., 1989b). A more detailed description of the health consequences of injecting drug use is given in Chapter 5 where we consider ways in which harm associated with opioid dependence might be assessed.

In this chapter we outline the basic facts about HIV and AIDS, and the epidemiology of HIV infection among injecting drug users. We then review the research to date on what is known about the sharing of injecting equipment and unprotected sex which are the specific risk behaviours associated with HIV transmission in this group. Finally, we consider the evidence for the effectiveness of methadone maintenance treatment in preventing HIV infection among injecting drug users and its further spread to other sectors of the community. As will be seen from this evidence, there is good reason to believe that methadone maintenance is the best available treatment option for preventing HIV infection among opioiddependent drug users. The implications of the advent of HIV and AIDS for methadone maintenance treatment are discussed in Chapter 11.

I. The Acquired Immune Deficiency Syndrome

Aids results from a person being infected with HIV. After being infected with HIV, the first clinical manifestation, known as seroconversion illness or acute HIV syndrome, usually occurs after a period of 3-6 weeks and consists of a relatively mild illness of brief duration, the symptoms of which include fever, sore throat, swollen lymph nodes, night sweats and headaches. This acute HIV syndrome is usually followed by a period during which few or no symptoms are observed. After this latency period, which varies between one and ten or more years, a progressive deterioration of immune functioning is observed which manifests itself in a variety of infections which are usually resisted by a healthy individual. As Sobel (1991) points out, it is too early in the course of the epidemic for the maximum length of the latency period to have been determined.

Although the clinical course is different in each case, some people may develop a series of relatively minor infections like oral thrush or candidiasis, and a range of other symptoms, typically fatigue and weight loss, as their immune system progressively loses its capacity to resist infection. A person is said to have AIDS when they manifest one of a number of serious illnesses such as pneumocystis pneumonia or Kaposi's sarcoma. Death usually results from a combination of such infections (for more detail about the clinical manifestations of AIDS among injecting drug users see Glasner & Kaslow 1990; McCutchan, 1990; Sobel, 1991).

AIDS first came to the attention of physicians in Los Angeles and New York in 1980, although it has now been established through retrospective searching of medical records and testing of stored sera that there were cases before 1980. Epidemiological research to date suggests that HIV has been present in human populations for some time, but that a number of societal and technological changes have created the conditions for its global epidemic spread. According to Grmek (1990), these changes include: the invention of the hypodermic syringe; the discovery of blood types early this century and the ensuing ability to transfuse blood; the increase in international travel afforded by developments in transport; and the liberalization of sexual life and drug use during the 1960s. At the time of writing, infection with HIV has been reported among injecting drug users in 30 countries throughout the world (Des Jarlais, 1992a).

The Epidemiology of HIV Among Injecting Drug Users

There are three main modes of transmission of HIV: through the introduction of HIV- infected blood or blood products into the body by practices such as transfusing blood products and injecting with HIV-contaminated needles and syringes; through the exchange of HIV- infected body fluids during unprotected sex; and by being passed from mother to child during pregnancy or birth, and perhaps via breast milk during feeding (Glasner & Kaslow, 1990). Certain practices such as sharing injection equipment and unprotected sexual intercourse have been shown to be important risk factors in the spread of HIV (Turner et al., 1989).

After homosexual men, injecting drug users constitute the second largest infected group in most Western countries, and in some countries (e.g. Italy and Spain) are the largest group (Friedman & Des Jarlais, 1991). As a group, they are potentially at risk for both sexual and parenteral HIV transmission. Their sexual partners are at risk if they practice unsafe sex, and children born of these relationships are also at risk for perinatal transmission. As will be seen below, there is ample evidence to suggest that injecting drug users are a potential source for HIV infection for the noninjecting drug-using population via sexual transmission of the virus.

Des Jarlais et al. (1989) have divided the introduction of HIV into a defined risk group into three phases: an introductory phase in which HIV is introduced and begins to spread; a rapid spreading phase during which the virus spreads rapidly among the group members (if facilitating conditions exist); and a stabilization phase during which the incidence of HIV infection levels out. In New York, for example, HIV was probably introduced among injecting drug users as early as 1975. After that, the incidence of HIV increased slowly until 1978 (introductory phase), when it spread rapidly among the group until 1981 (rapid spreading phase), after which the rates of HIV infection leveled out. During the rapid spreading phase, the percentage of infected drug users went from approximately 20% to 50% (Des Jarlais et al., 1989). The seroprevalence rate has since remained at around 50-60%. Similar rates of spread among injecting drug users have been observed in many other areas throughout the world (see Des Jarlais, 1992a). Des Jarlais et al. (1989) have argued that it is important that attempts to prevent the spread of HIV be put in place before the onset of this rapid spreading phase, because it appears that once infection reaches a critical level in a given population of drug injectors, and given the right conditions, it only takes a short period of time for it to reach epidemic proportions.

According to Des Jarlais (1992a), two conditions that allow for the rapid spread of HIV are: a) low levels of appreciation of the risks associated with AIDS and injecting behaviour; and b) frequent sharing of injecting equipment with multiple sharing partners, such as occurs in 'shooting galleries' and the houses of drug dealers who provide injecting equipment at the time of purchase. Friedman and Des Jarlais (1991) have noted that those areas where HIV has spread most rapidly are those where heroin is the major drug of injection, though they also note that this may simply be due to a lack of information about the epidemiology of HIV among injecting drug users in countries where cocaine is the major drug of injection.

Rapid spread of HIV is by no means inevitable. In some cities, infection rates appear to have stabilized at much lower levels (e.g. San Francisco where it remains at around 12%). Friedman and Des Jarlais (1991) suggest a number of reasons why these differences have occurred: the differential effectiveness of HIV prevention efforts; the extent to which injecting drug users have changed their risk behaviours; the time at which HIV was introduced to a given population of drug users and, more critically, its relationship to the discovery of the transmission characteristics of the virus and the ensuing risk reduction campaigns; and local variations in the rituals and lifestyle associated with drug injecting.

Socio-economic factors and ethnicity are important in any characterization of the HIV epidemic among injecting drug users. In the USA, for example, black and Hispanic users have far higher rates of HIV infection than do white drug users (Friedland, 1989). The association between membership in these two groups and socioeconomic deprivation is important. As Friedman and Des Jarlais (1991) observe, socioeconomic disadvantage is not only a discernible risk factor for HIV infection but for injecting drug use as well. Ultimately, reducing injecting drug use, and therefore risk of HIV infection, among disadvantaged populations requires larger scale interventions than drug treatment programs can offer.

In Australia, HIV infection among injecting drug users has been surveyed in a number of ways. Wodak et al. (1987) in a study to determine the presence of HIV antibodies in needles and syringes exchanged in Sydney in late 1986 and early 1987 found that 1% contained infected blood. In a similar study conducted in the same city during July and August 1987, 3% of the syringes returned were found to be contaminated (Wolk et al., 1988). More recent estimates suggest that around 3% to 6% of injecting drug users in urban centres in Australia are HIV seropositive (Darke et al., 1992; Morlet et al., 1990; Ross et al., 1991). Finally, 4% of reported AIDS cases in Australia have injecting drug use identified as a risk factor (National Centre on HIV Epidemiology and Clinical Research, 1991).

There is as yet no cure for AIDS. The only means available to deal with HIV, therefore, is by eliminating or modifying the risk behaviours involved in its transmission (Becker & Joseph, 1988; Kelly & Murphy, 1991). In the case of injecting drug use, these changes amount either to abstinence or, if that goal is not feasible, to either changing the route of ingestion of the drug (say to smoking or swallowing) or to safer injecting, which means not sharing injecting equipment or properly cleaning this equipment if the user continues to share. As Becker and Joseph (1988) have pointed out, abstinence for life might seem an attractive option, but encouraging modification of injecting practices is more likely to be effective for many drug users. They argue that the consequences of not adopting measures that take into account likely behaviour change could be disastrous. It is possible that a limited number of quite active, infected individuals could maintain and perhaps increase the incidence of infection among injecting drug users.

Methadone maintenance is an intervention that takes into account what is achievable with this population. It provides an orally ingested alternative to injected opioids such as heroin, which makes it an attractive option for opioid users who are not prepared, or who are unable, to become abstinent. This means that a reduction is achieved in the risks associated with the use of opioids without requiring abstinence on the part of the drug user. Other advantages are that methadone maintenance programs can also function as dissemination points for education about HIV and AIDS and provide contact with the health care system for those injecting drug users who are already infected.

(Top)

II. HIV Risk Behaviours:
Injecting and Needle Sharing

II. a. Risk Factors Associated With HIV Infection
II. b. The Dynamics of Needle Sharing
II. c. Characteristics of Needle Sharers

HIV is transmitted via injecting drug use through the sharing of injection equipment that has been contaminated with HIV-infected blood. As we have already stated, needle sharing has been and remains a common practice among injecting drug users, albeit at a lower prevalence than was the case before HIV. With the advent of HIV and the ensuing public health campaigns to reduce HIV risk-taking behaviours, reductions in needle sharing have been reported in both the USA and in Europe (e.g. Des Jarlais et al., 1989; Durante, 1991; Harris et al., 1990; Kall & Olin, 1990; Klee et al., 1990a; Power et al., 1988; Ronald et al., 1992; Selwyn et al., 1987). Contrary to popular images of being irresponsible, out-of-control individuals, injecting drug users have responded positively to AIDS awareness and risk reduction campaigns (Friedland, 1989). However, as the evidence below shows, some drug users continue to share needles and therefore continue to expose themselves and others to the risk of being infected. In this section, we consider under what circumstances injecting risk behaviours occur and what, if anything, can be said about those users who continue to engage in these risk behaviours.

In Australia, studies that have surveyed the needle sharing practices of injecting drug users have found it to be a common practice. Morlet et al. (1990), in a survey of users attending the Albion Street (AIDS) Centre in Sydney, found that 87% of the sample had shared injecting equipment at some time in their lives. Wolk et al. (1990a) found similar rates, with 80% having shared in their current drug use period and 74% of these having shared with more than one person. Darke et al. (1990) report a rate of 40% having shared in the month prior to interview and 24% of these having shared with more than one other person. These figures are consistent with prevalence studies of hepatitis B and C among this population, where high rates (89% and 86% respectively) suggest that most injecting drug users have shared needles at some time in the past (Bell et al., 1990c).

If drug users do share needles, then the cleansing procedures they use are an important determinant of the probability of transmitting HIV. Rinsing the injecting equipment with bleach has been the most commonly recommended decontamination procedure if it is necessary to share at all (e.g. see Newmeyer, 1988). It is disturbing then to note that Wolk et al. (1990a) found that even though 86% of the needle sharers cleaned used needles before injecting with them, 80% of this group rinsed at some time with water, 8% with alcohol, and only 30% boiled the equipment. The authors note that the interviews were conducted before the campaign to educate users about the need to rinse shared needles with bleach. However, the Darke et al. (1990) study, which was conducted after this campaign, found that even though 89% of sharers rinsed before injecting, only 15% of this group had used bleach. Darke (in press), in a review of the literature on HIV and injecting drug use, has found that the low levels of cleaning with bleach found in Australia are similar to findings in the USA (with the exception of San Francisco) and Europe.

II. a. Risk Factors Associated With HIV Infection

A number of factors have been associated with an increased risk for contracting HIV through needle sharing. The frequency with which the user injects has been found to be associated with HIV seropositivity (e.g. Chu, et al., 1989; Schoenbaum et al., 1989). Duration of injecting drug use has also been associated with being HIV positive (Chu et al., 1989; Schoenbaum et al., 1989), although this presumably reflects the relationship between the time of beginning injecting and the advent of risk reduction campaigns among injecting drug users. Related to frequency of injecting is the issue of poly-drug use. Dolan et al. (1987), in an American study, found that drug users who used multiple drugs were more likely to ,share, as did Klee et al. (1990b) in an English study. Darke et al. (1990) also found that poly-drug use was associated with greater needle sharing in their study of Sydney injecting drug users. One drug that has been associated with more frequent injecting and needle sharing is cocaine (Chaisson et al., 1989; Darke et al., 1992; Schoenbaum et al.' 1989; Torrens et al., 1991). Cocaine has a much shorter elimination half-life than heroin, meaning that it has to be injected more frequently to maintain its effects. A history of cocaine abuse has been found to be associated with being HIV positive in the United States (Chaisson et al. 1989; Schoenbaum et al., 1989).

II. b. The Dynamics of Needle Sharing

Injecting drug users share injection equipment for a variety of reasons. Not being in possession of a new needle and syringe at a time when illicit drugs are available leads users to either share their equipment with each other, or to use old equipment that they already have in their possession (Klee et al., 1990a; McKeganey et al., 1989; Selwyn et al., 1987). Users who report greater difficulty in tolerating withdrawal symptoms are also more likely to share (Magura et al., 1989a). This group of users appears to be willing to inject with used equipment rather than tolerate the period of withdrawal necessary until clean equipment can be obtained. As might be expected, the easy availability of clean injecting equipment has been found to be associated with reductions in needle sharing (e.g. Calsyn et al., 1991a).

A number of social factors have also been found to increase the incidence of needle sharing. Having a partner or friends who inject has been a factor associated with needle sharing (Klee et al., 1990a; Magura et al., 1989a). A number of important features of this variable have also been isolated. Sharing needles with family members, sexual partners and friends has been shown to function as a sign of bonding (Magura et al., 1989a; McKeganey et al., 1989). With sexual partners, who often do not regard their sharing as 'true' sharing, injecting each other may be an expression of intimacy (Turner et al., 1989). Peer pressure is an important component of the renewal of family and friendship bonds and has been reported to contribute to continued needle sharing (Magura et al., 1989a). Peer pressure is also reported to be felt by some users in the form of a duty to help other users who need their equipment to inject (McKeganey et al., 1989). In a study of the impact of needle exchanges on sharing in north west England, Klee et al. (1991a) found that injectors who did not use the exchanges put pressure on those who did to share their used equipment.

New users and irregular injectors are two groups who are also at risk because clean injecting equipment is usually unavailable at the time of use even in areas where such equipment is available through needle and syringe exchanges. New users are usually initiated into injecting by experienced users who teach the injection skills necessary to successfully inject oneself (Turner et al., 1989). Usually, the skilled injector will inject the new initiate. Lacking their own equipment, new injectors are dependent upon the equipment of experienced users. Irregular injectors are unlikely to have equipment in their possession because their drug use is often unplanned and dependent on circumstance. In a study of drug users who use needle exchanges and those who do not, Hartgers et al., (1989) found that irregular drug use was associated with not using needle exchanges and with more needle sharing. This is consistent with the evidence above concerning the unavailability of injection equipment as a reason for sharing. In the Wolk et al. (1990a) study, 19% of those who shared said they did so because of an unplanned decision to use drugs. As well as these two kinds of users, Klee et al. (1990a) found that homeless drug injectors tended to share more than other drug users, a finding that has been reported elsewhere (Siegal et al., 1991).

II. c. Characteristics of Needle Sharers

There is growing evidence that a sub-group of drug users continue to share injecting equipment even though they know the risks associated with it (Selwyn et al., 1987; Wolk et al., 1990a). Some research has tried to determine if there are any definable characteristics of this group. Dolan et al. (1987) interviewed 224 males to identify variables that discriminated between sharers and non-sharers. They found that those subjects who had more severe drug problems and who used more drugs were more likely to share needles, even though they knew the risks involved. They rejected the notion that sharers had a particular personality profile or any specifiable psychopathology because of an absence of an association between sharing and subjects' scores on the Minnesota Multiphasic Personality Inventory, an instrument that is used for measuring personality variables.

Studies using more specific diagnostic criteria have, however, suggested otherwise. Nolimal and Crowley (1989) and Brooner et al. (1990) found that users with a diagnosis of antisocial personality disorder, a disorder frequently associated with illicit drug use, were more likely to share than other users. This finding suggests that users who fit into this diagnostic category may be less responsive to AIDS education aimed at changing behaviour when it is considered along with other evidence that has found that clients with antisocial personality disorder do not respond well to treatment (Brooner et al., 1990). Metzger et al. (1991) have also found psychiatric diagnosis to be associated with needle sharing, with methadone maintenance patients who were more depressed or who had more severe psychiatric problems overall being more likely to share. Antisocial personality disorder and other co-morbid psychiatric states are discussed in more detail in Chapters 8 and 13.

One other factor that has been found to be related to needle sharing is the concomitant use of benzodiazepines. In a study of HIV risk behaviour recently conducted in the north west of England, it was found that the tranquilizer temazepam was associated with needle sharing (Klee et al., 1990b). Temazepam users used more types of drugs, shared more often and had shared more recently than other drug users, including those who used tranquilizers other than temazepam. Temazepam users also had less friends and were involved in criminal activity to a greater extent. This association between the use of benzodiazepines has been found in other studies (Darke et al., forthcoming; Metzger et al., 1991).

III. HIV Risk Behaviours: Sex

The sexual transmission of HIV takes place through the exchange of body fluids that occurs during unprotected penetrative sex - that is, penile penetration of the vagina or anus (and perhaps the mouth if ejaculation takes place) without a condom. In the general population, the type and number of unsafe sexual contacts is known to be important in the spread of HIV (Turner et al., 1989). Injecting drug users who engage in needle-sharing behaviours are an important risk factor for their sexual partners. This risk of heterosexual transmission was identified early on in the epidemic (Sobel, 1991) and confirmed in findings like those of Schoenbaum et al. (1989) who found in a New York study that heterosexual contact with an injecting drug user was a risk factor for being infected with HIV.

Almost nothing is known about the sexual behaviour of injecting drug users as a group. This reflects a similar state of affairs concerning the general population (Turner et al., 1989). The frequency of specific types of sexual behaviours and the number and frequency of sexual contacts for most definable social groups is unknown. This lack of knowledge makes the prediction of the spread of HIV difficult. What is known to date concerning the sexual behaviour of drug users and the risk for HIV is reviewed below, outlining first the risk factors associated with sexual behaviour and then factors that have been found to be associated with increased or continued risk taking.

III. a. The Sexual Behaviour of Injecting Drug Users

Injecting drug users have been found consistently to be a sexually active group in the USA, Europe and Australia (Darke et al., 1990; Donoghoe et al., 1989; Feucht et al., 1990; Klee et al. 1990c; van den Hoek et al., 1990; Wolk et al., 1990a). Sexually active injecting drug users report high levels of unsafe sex, considerable casual sex, and high rates of sexually transmitted diseases (STDs). Focusing on the Australian evidence, 69% of the sexually active subjects surveyed by Wolk et al. (1990a) reported very low or no use of condoms. Many of the subjects had had more than one sexual partner in the recent past. This finding is consistent with overseas findings concerning low condom usage and multiple sexual partners among many injecting drug users (e.g. Donoghoe et al., 1989; Durante, 1991; Lewis et al., 1990; Magura et al., 1990a).

III. b. Prostitution

Prostitution is a common way for female, and to a lesser extent male, drug users to support their drug use. Wolk et al. (1990a) found that 13% of males and 41% of females had worked as prostitutes since 1981, and 13% of the subjects interviewed were currently engaged in prostitution. Ten per cent of the Darke et al. (1990) sample and 15.5% in the Morlet et al. (1990) study reported current prostitution. Because prostitutes tend to have high numbers of sexual contacts, the use of condoms by this group is of interest. Overall the evidence suggests that sex industry workers have responded to the advent of HIV, and groups organised by sex workers themselves have been important in the widespread adoption of safe sex practices at work (see Donoghoe, 1992). It has been found, however, that drug injecting prostitutes use condoms at work but not with their regular sexual partners (Darke et al., 1990; Klee, et al., 1990a; Turner et al., 1989; van den Hoek et al., 1989; Wolk et al., 1990a). Van den Hoek et al. (1989) pointed out, however, that in their study most of these private sexual partners tended to be injecting drug users themselves, which may constitute a high risk if either party has been sharing injecting equipment.

Two findings suggest that prostitution may still be of concern. Bellis (1990) found low condom use by drug injecting prostitutes who worked the streets in five southern Californian cities. None of the subjects in this study was in any form of drug treatment. It may be that it is those prostitutes who are not in any form of drug treatment, and who work at the lower end of the market (e.g. on street corners), who may be under more pressure to engage in unsafe sex. The other finding of concern is that of Van den Hoek et al. (1989) who, in their study of drug injecting prostitutes in Amsterdam, found that although the group as a whole reported reasonable use of condoms at work, 81% had reported contracting STD in the six months prior to interview. This anomaly suggests that self-reported rates of unsafe sex may not always be accurate. The main risk factor for HIV infection among the prostitutes in their study remained injecting drug use.

III. c. Other Sexually Transmitted Diseases
Among Injecting Drug Users

A high incidence of sexually transmitted diseases (STDs) within a group is a strong indicator that the group is engaging in high levels of unsafe sex (van den Hoek et al., 1990). Injecting drug users in Sydney have high rates of reported infection with STDs (Morlet et al., 1990; Ross et al., 1991). Half of the Morlet et al. (1990) sample and just over one-third of the males and one-half of the females in the Ross et al. (1991) study reported having an STD at some time in their life. Besides indicating high levels of unsafe sex, these findings are of concern because ulcerative genital lesions are thought to heighten the risk of HIV transmission during sexual intercourse and have been found to be associated with HIV seropositivity among injecting drug users (Morlet et al., 1990; Trapido et al., 1990; van den Hoek et al., 1989).

III. d. Homosexual and Bisexual Males

Homosexual and bisexual male injecting drug users who engage in unsafe sex are at greater risk for infection with HIV. They have higher rates of HIV infection and STDs than male heterosexual and female drug users who have approximately equal rates (Morlet et al., 1990; Ross et al., 1991). The differential rates of infection for HIV for each of these groups are similar to those found in the general population. Homosexual men tend to have higher rates of infection with HIV than bisexual men, who in turn have higher rates than heterosexual men (Ross et al., 1991).

III. e. Factors Associated With Unsafe Sex

Some factors have been found to be associated with an increase in unsafe sex. Darke et al. (1990) found that younger injecting drug users tend to engage in more unsafe sex than older users. Both Klee et al. (1990a) in the United Kingdom and Loxley et al., (1991) in Western Australia have found that younger users also tend to have more casual sex. As with needle sharing, the use of temazepam (Klee et al., 1990b), antisocial personality disorder (Brooner et al., 1990; Nolimal & Crowley, 1989) and being homeless (Klee et al., 1990a) are associated with having more sexual partners and not using condoms. Partner pressure has been found to be another important factor in condom use. The extent to which a patient perceives their partner to be receptive to suggestions about safe sex has been found to be related to level of condom use (Magura et al., 1990).

Although injecting drug users are, at this point in time, primarily at risk because of their injection practices, they do provide the most important transmission bridge to the heterosexual population at large. In the USA, just over 60% of AIDS cases reported in women for the two years April 1987 to March 1988 and April 1988 to 1989 had heterosexual contact with an injecting drug user as the suspected route of HIV transmission (Feucht et al., 1990). In New York, which has the highest concentration of injecting drug users in the USA, sexual contact with an injector was the source of 90% of cases of heterosexually transmitted AIDS and of 80% of the cases of maternal transmission (New York City Department of Health, 1988 cited in Des Jarlais et al., 1990). Many injecting drug users have non-injecting sexual partners and, given the above evidence about high rates of unsafe sex, could become the maijor source for heterosexual transmission in other parts of the world (Abdul-Quader et al., 1987; Donoghoe et al., 1989; Klee et al, 1990c). Reports concerning the differences in risk for males and females with regard to heterosexual transmission of HIV are misleading when it comes to prevention. The risk to males when compared with females would have to be inconceivably low before such differences would matter (Kaslow & Francis, 1989).

It is a common finding that drug treatment and education campaigns are more effective at reducing needle sharing than they are at reducing unsafe sex (Donoghoe, 1992). The continued practice of unsafe sex by injecting drug users in spite of education campaigns and contact with drug treatment programs presents a challenge for methadone maintenance programs to develop more effective ways of modifying risky sexual behaviour. This challenge is important not only for the patients in methadone treatment but also for their sexual partners and their children.

IV. Methadone Maintenance and Containment of HIV

As is evident from the previous two chapters of this book, methadone maintenance treatment is effective in reducing illicit opioid use and therefore is also likely to be effective in reducing those behaviours associated with injecting opioid use that are involved in HIV transmission. This substantial body of evidence has to be considered as part of the case for the role of methadone maintenance treatment in HIV prevention. For the purposes of this chapter, the focus is on the impact of methadone maintenance on the frequency of injecting and sharing. As already stated above, methadone maintenance treatment has not as yet had any substantial effect on the levels of unsafe sex among its patients. Ways in which methadone treatment might contribute more in this regard are considered in Chapter 11.

IV. a. Methadone Maintenance and HIV Infection

One indicator of the success of methadone treatment in protecting its patients from infection with HIV is whether it has protected them from infection in places where HIV has already spread rapidly among injecting drug users. As already noted, Des Jarlais et at. (1989) have observed that prevention attempts are most likely to be effective during the rapid spreading phase of HIV infection. Retrospective studies have found an association between length of time in methadone treatment and low rates of seropositivity. Patients who entered methadone treatment in New York before 1982 were found to be less likely to be HIV positive than those who had entered treatment after that year (Abdul-Quader et al., 1987). Schoenbaum et al. (1989), again for injecting drug users in New York, found that there was an inverse relation between total months of methadone treatment since January 1978 and the presence of HIV antibodies.

In a group of long-term, stable patients who entered treatment before the spread of HIV in New York, Novick et al. (1990) reported that none was seropositive, even though 91% had been exposed to hepatitis B, indicating that nearly all of them had shared needles at some time. Two other studies have found that patients in methadone treatment were less likely to be HIV positive than those in detoxification programs (Marmor et al., 1987) and those not yet receiving methadone (Chaisson et al., 1989). Similar findings suggestive of the effectiveness of methadone maintenance in preventing HIV infection among its patients have been reported for Italy and Sweden (see Des Jarlais, 1992a).

Although the studies discussed in the previous paragraph are consistent with the evidence for the overall effectiveness of methadone treatment, there are other plausible explanations for many of these findings (Ward et al., 1992). It could be the case, for example, that those patients who remain in methadone maintenance are less likely to engage in risk behaviours than either patients who leave methadone treatment early or injecting drug users who do not enter treatment. The finding of lower rates of HIV seropositivity among methadone patients would then be due to naturally occurring selection between those who enter treatment and those who do not.

The Swedish study (Blix & Gr6nbladh, 1988; described also by Des Jarlais, 1992b) is of special interest in this regard, because the way in which patients were accepted into methadone treatment approximated a random selection procedure. The important point to be made about this study is that nearly all the patients who entered treatment after 1983 had previously applied and been refused. Applicants for methadone maintenance were accepted or refused treatment in an almost random fashion which depended on whether there was a place available at the time of application or not. Three per cent of patients who entered methadone maintenance before 1983 were found to be HIV positive, compared with 16% of those who entered treatment during the years 1984 to 1986 and 57% of patients who entered treatment after 1987. There had been no seroconversions of any patients who had tested negative for HIV antibodies on entry since 1984. This study provides stronger evidence in that it demonstrates that methadone maintenance protects its recipients from HIV infection and that this appears to be independent of selection bias. The following section reviews evidence that methadone maintenance treatment specifically affects risk behaviours associated with injecting. There is still, however, a need for better controlled prospective studies to confirm the findings that methadone maintenance protects its recipients from HIV infection.

IV. b. The Effectiveness of Methadone Treatment in Reducing Injecting and Needle Sharing

The largest study to date of whether methadone treatment reduces injecting and needle sharing has been the Three Cities Study conducted by Ball and his colleagues, which is reviewed in detail in Chapter 3 (Ball & Ross, 1991; Ball et al., 1988). The outcomes relevant to HIV were the number of subjects that reported injecting and the number of days per month that they had injected. For sharing, the same measures were employed - the number of subjects sharing and the number of sharing days as an indicator of frequency of sharing.

For injecting, methadone treatment had a marked effect on both injecting and on the frequency of injecting for those who did inject. Of the 388 subjects who had remained in methadone maintenance until the end of the study period, 36% had not injected again after one month of treatment. A further 22% had not injected in the past year, and a further 13% had not injected for a period of between one and 11 months. The remaining 29% had injected in the last month. Overall 71% had not injected in the month prior to being interviewed. These data are consistent with the notion that methadone treatment leads to a slow reduction in injecting and its eventual cessation for most patients while they remain in treatment.

Similar results were found for needle sharing - both the number of sharers and the frequency with which they shared were reduced. Those who shared during their last period of injecting shared less than those whose last period of drug use occurred before or during the admission phase of treatment. Twenty-nine of the group who had injected in the previous month (9% of the overall sample in treatment) had shared needles.

These results are supported by a number of other findings. Selwyn et al. (1987) found that being in methadone treatment was associated with a decrease in both needle sharing and injecting drug use. The subjects in the Abdul-Quader et al. study (1987) who had been in methadone treatment longest had the lowest levels of these risk-taking behaviours. In their Sydney study, Darke et al. (1990) found that of the 20% of their subjects in treatment (the majority of whom were in methadone treatment) reported needle sharing in the month prior to interview compared with 68% of the subjects who were not in treatment. In this case it was unlikely that drug users in treatment were less likely to report such behaviour, as it was demonstrated through collateral interviews with subjects' sexual partners that patients in treatment were reasonably truthful about their risk-taking behaviour when interviewed by independent researchers (Darke et al., 1991). Klee et al. (1991a) report similar findings in an English study, finding that long-term methadone maintenance was associated with a reduction in needle sharing, but only for older patients who had been long-term injecting drug users. Younger patients had the same levels of risk taking as patients out of treatment. Finally, in terms of injecting safely, Hartgers et al. (1989), in a Dutch study, found that needle exchange users tended to have had more contact with methadone maintenance programs over the previous five years than did injectors who did not exchange, suggesting that methadone maintenance programs may be effective as dissemination points for information about safe injecting practices.

Although all of the studies reviewed here have been observational studies without control groups, the evidence is reasonably consistent that methadone treatment is effective in reducing the HIV risk behaviours associated with injecting. This consistent finding - supported by the evidence reviewed in the previous two chapters - that methadone treatment is effective at reducing injecting drug use is difficult to ignore. In the absence of evidence to the contrary, it is reasonable to conclude that methadone maintenance is an important HIV prevention measure among opioid- dependent injecting drug users.

IV. c. Are Some Methadone Programs More Successful Than Others?

As we have observed in the previous chapters, much of the research concerned with the effectiveness of methadone treatment has focused on the issue of whether the treatment works at all. Methadone maintenance programs, however, differ from each other in substantially important ways, and very little research has looked at what types of programs are the most successful (Ball & Ross, 1991). Knowledge of this sort is important to the issue of HIV containment so that methadone maintenance programs can respond in the most effective way possible. Two aspects of treatment that are important in delivering effective treatment are dosage and the duration of methadone maintenance, both of which are dealt with in detail in Chapters 6 and 9 respectively. It is worth noting briefly that two studies have looked specifically at dosage in relation to HIV. Brown et al. (1989a), in a study of 454 methadone patients in New York, found that low methadone doses were associated with being HIV positive. This held true even after total time spent in drug treatment was controlled for a variable that was also predictive of HIV status. In the Ball et al. study (Ball & Ross, 1991; Ball et al., 1988), patients on lower methadone doses were more likely to be currently injecting. These findings suggest that lower doses of methadone (<60 mg) may be less effective than higher doses in preventing HIV infection, a conclusion that is consistent with the evidence reviewed in Chapter 6. Although by themselves these two studies do not provide strong evidence for a causal relationship between methadone dose and HIV infection, taken together with the evidence in Chapter 6 they are strongly suggestive of one.

IV. d. Cocaine

As mentioned above, methadone treatment is primarily a treatment for opioid dependence. The treatment situation for methadone maintenance programs has been complicated recently in the USA by the advent of the widespread use of cocaine, and to a lesser extent in Australia by the use of amphetamines. Methadone does not provide cross-tolerance for cocaine and would not be a recommended treatment for someone with primary cocaine dependence. Two studies have tried to separate the effects of methadone on injecting injecting heroin and injecting other drugs. Magura et al. (1989a) found that of 110 methadone patients who continue to inject, very few were injecting heroin leading them to conclude that methadone maintainance was effective in eliminating heroin use. Chaisson et al. (1989) also found an association between being in methadone treatment and a substantial reduction or cessation of heroin use. They also found methadone treatment was associated with a reduction in cocaine injection for more than half the methadone patients. Cocaine injection, however, remained a major problem for these programs. In Australia, a recent increase in the availability of amphetamines is comparable to that seen with cocaine in the USA. The extent to which amphetamine use in Australia is associated with HIV infection has yet to be established, although some research suggests that Australian amphetamine injectors were less likely to engage in risky behaviour than heroin injectors (Hall et at.,in press) .

IV. e. The Fate of Methadone Treatment Drop-outs

While injecting decreases over time for most patients in methadone maintenance programs, leaving a program is associated with an increase in injecting drug use. In the Ball et al. study (Ball & Ross, 1991), the subjects who had dropped out of treatment during the year to follow- up interview were much worse off than their counterparts who remained in treatment. By the time of the interview 68% of drop-outs had relapsed to injecting drug use,.with 27% of the relapsers having shared needles. Although it could be argued that these figures are inflated because treatment drop-outs differ from other patients, this does not appear to be the case in the Ball et al. study. The patients who dropped out of treatment only differed on three of the characteristics assessed: they had received less methadone maintenance treatment overall; more were unemployed; and they were slightly younger than those that stayed in treatment. However, although the drop-outs did not differ in terms of the aspects of their drug dependence on which they were assessed (age of onset , length of dependence etc.), data were not provided that compared rates of injecting among the drop-out group while they were in treatment with those of the patients who remained in treatment. Murphy and Rosenbaum (1988) also claim that clients who were mandatorily withdrawn from methadone after two years because of a change in treatment policy were at greater risk for HIV infection due to relapse, but they present no data to support these conclusions. Findings like these are difficult to ignore and have important implications for treatment duration and dismissal of patients from programs. Uchtenhagen (1990a), for example, suggests that expectations about patients and the regulations of methadone maintenance programs should be relaxed in an attempt to keep patients in treatment.

IV. f. Low Threshold, Low Intervention and Interim Methadone Programs

Increasing the effectiveness and the attractiveness of methadone treatment may necessitate offering different kinds of services and trying to reach a wider range of opioid users, and accordingly a number of innovative programs have been proposed and implemented (Uchtenhagen, 1990a). Low intervention methadone maintenance programs are those in which methadone is dispensed with minimal ancillary services. Such programs may or may not be 'low threshold', implying that there are few restrictions in terms of entry criteria. Dispensing low doses of methadone by bus in The Netherlands is another innovation (Buning et al., 1990). The interim maintenance program for people on waiting lists for methadone treatment discussed in Chapter 2, which existed for a brief time in New York, is another variant of these new post-HIV forms of methadone maintenance (Yancovitz et al., 1991).

These new forms of methadone treatment have a number of rationales. One is that making treatment easier to access will mean that more injecting drug users at risk will be treated. Another is that treatment might be more attractive to drug users if less therapeutic demands are made and might therefore encourage subsequent entry to more change-oriented forms of treatment. Such programs also have an economic attraction for governments. Low intervention methadone maintenance programs are a much cheaper option than the more intensive variants of methadone treatment, and a broad move to low threshold treatment would mean that more patients could be maintained on methadone for the same cost.

Schuster (cited in Nathan & Karan, 1989), however, has warned that a danger associated with low intervention programs is that they may be seen as a cheap alternative to traditional methadone treatment if this means compromising the quality of care being delivered. This point is important, because these new variants of methadone treatment have yet to be properly evaluated and, although there is reason to believe that they may reduce injecting drug use, there is little direct evidence at present that they are effective.

V. Summary

The advent of HIV has meant a new life threatening addition to the existing risks associated with injecting drug use. Evidence from New York and other cities throughout the world suggests that once HIV is introduced to the injecting drug using population it can spread rapidly. The main mode of transmission that initiates and maintains this spread among drug injectors is the sharing of contaminated injecting equipment. HIV is also spread sexually and perinatally, thus putting at risk the sexual partners and children of injecting drug users. There is as yet no cure for HIV infection, so the only way in which the pandemic will be stemmed will be through the permanent modification of behaviours known to be routes of transmission.

Some of the dynamics of needle sharing have been identified. The frequency with which a person injects, the type of drugs they use, and having a long history of injecting drug use have all been found to be significantly associated with being HIV positive. The sharing of injecting equipment is influenced by peer pressure, having family and friends who are users, the absence of easily available injecting equipment, being unable to tolerate withdrawal symptoms, injecting irregularly, being homeless, having a diagnosis of antisocial personality disorder, and the use of benzodiazepines.

HIV is transmitted sexually via the exchange of body fluids that occurs during unprotected penetrative sexual intercourse. This mode of transmission in combination with the sharing of injecting equipment means that the sexual partners of injecting drug users are also placed in a high risk category. Little is known about the sexual behaviour of the general population or the population of injecting drug users which makes predicting the spread of HIV difficult. The little research available suggests that drug injectors are a sexually active group and that the incidence of unsafe sex among this group is quite high. This evidence is confirmed by studies which have shown that having a sexual partner who injects is a discernible risk factor for HIV infection. The high incidence of STDs among injecting drug users suggests that unsafe sex is common among this group. Research to date suggests that drug users who work as prostitutes engage in safe sexual practices at work but do not do so when not working. Homosexual and bisexual male injecting drug users are at greater risk if they engage in unsafe sex due to the higher rates of infection among these groups in general. Unsafe sex among injecting drug users has been found to be associated with being younger, with using benzodiazepines, with being homeless, with having antisocial personality disorder, and with being unable to resist partner pressure about the use of condoms. As yet, methadone treatment has had little influence on the unsafe sexual practices of its patients.

Being in methadone treatment has been associated with lower rates of HIV infection for opiate injectors than not being in treatment in cities where the incidence of infection is quite high. Evidence also suggests that methadone treatment is effective at reducing injecting drug use and needle sharing. Successful methadone programs retain their patients in treatment, have higher maintenance dosages, have low staff turnover, and develop close, long-term relationships with their patients. Patients who drop out of methadone treatment have been found to have higher levels of HIV risk-taking behaviour than those who remain in treatment. Attracting HIV positive injecting drug users into methadone treatment is also important. Low threshold methadone programs may be a way of widening the appeal of methadone treatment, thereby reaching members of the injecting drug using population who would not otherwise consider treatment as an option.