Prevalence of Psychiatric Disorder in a Methadone Maintenance Population

Abstract

Objective: The objective of this study was to examine the prevalence of psychiatric disorders in a group of patients who had recently entered a methadone maintenance programme.

Method: A total of 62 patients were interviewed using the Composite International Diagnostic Interview (CIDI) within 6 months of commencing methadone maintenance. The CIDI was used to establish symptoms of psychiatric illness at interview and in the 12 months prior.

Results: In the 12 months prior to interview, 76% of the sample fulfilled ICD-10 criteria for a psychiatric disorder other than substance-use disorder. Over half of the group interviewed fulfilled ICD-10 criteria for an affective disorder, two-thirds fulfilled criteria for an anxiety disorder and just under half fulfilled diagnostic criteria for both an affective disorder and an anxiety disorder in the 12 months prior to interview. At the time of interview, 19% fulfilled ICD-10 diagnostic criteria for a moderate or severe affective disorder. Seventy per cent of males and 89% of females interviewed had a comorbid psychiatric illness. In 71% of the group who had a comorbid psychiatric illness, the onset of psychiatric symptomatology was reported to predate the use of heroin.

Conclusion: The prevalence of psychiatric disorder is up to 10 times higher in the population on methadone maintenance than in the general population and is two to three times higher than that found in community surveys of those with a substance-use disorder. These results are consistent with earlier findings and have implications for service planning.

The recent increase in the use of heroin in Australia is a major cause of alarm to the community. A 1998 survey of Victorian households showed that almost 1% of the population aged 14 years and over had used heroin in the previous 12 months compared with 0.2% in 1995 [1]. The price of heroin in Victoria has fallen from approximately $600 for a gram in late 1997 to $300 a gram in mid-1998 while the number of heroin-related deaths has risen from approximately 50 in 1991 to 360 in 1999 [1].

It is widely recognized that those who have a substanceuse disorder also have an increased risk of other psychiatric disorders [2–6].

Rounsaville et al. [7] used the Schedule for Affective Disorders and Schizophrenia – Lifetime version [8], and found that in a group of 533 opiate-dependent individuals, 74.3% met criteria for some affective disorderand 26.5% for antisocial personality. Milby et al. [9] studied a sample of 102 individuals who had been in a methadone programme and found that at least one anxiety disorder was diagnosed in 55%, an affective disorder was diagnosed in 58% and an anxiety disorder coexisted with an affective disorder in 36% of the sample. In a study of depression among 528 needle exchange programme and methadone maintenance participants, Brienza et al. [10] found that 54% of the former and 42% of the latter met criteria for major depression.

In an Australian study, Darke et al. [11] usedthe 28-item version of the General Health Questionnaire [12] to measure levels of psychopathology in a group of 239 opioid users (including 210 subjects who were enrolled in a methadone maintenance programme at the time of the study). They found that 59% showed evidence of significant psychopathology.

The prevalence of coexisting psychiatric disorders in those with substance-use disorder is higher in females than males [6] and this increased prevalence has also been found in those who use opioids [11].

The prevalence of substance use is changing nationally with changes in illicit drug availability and price and with changes in drug culture preference [1]. There are also likely to be differences in the prevalence of drug use between settings, such as rural and urban. Consequently, studies of prevalence are needed in a variety of settings in order to draw valid conclusions with which to inform needs analysis and service planning. This study was undertaken at the instigation of the three general practitioners who provide services to the Barwon methadone maintenance programme and who wished to identify those patients who are commencing methadone maintenance therapy and have other treatable psychiatric illness. An additional aim of this study was to examine the prevalence of psychiatric disorders in a group of patients presenting to a methadone maintenance programme in an urban/rural regional setting.

Method

Barwon Health: Community and Mental Health provides mental health and drug treatment services to a population ofapproximately 230 000 people living in south-western Victoria where about 50% of the population live in the city of Geelong and the remainder in the surrounding rural area. The methadone maintenance programme operates at the Drug Treatment Services clinic, which is situated within a Community Health Centre in central Geelong. Referrals to the programme are from local general practitioners and methadone maintenance therapy is provided to over 300 patients. The doses of methadone prescribed range from 30 mg to 100 mg daily per patient. Drug treatment counsellors employed in the Drug Treatment Services clinic were asked to recruit subjects for the study over a 4-month period between July and October 1999. They invited patients to join the study if they presented to commence methadone maintenance therapy in that period or if they had been on methadone maintenance for less than 6 months. Thus all patients assessed as suitable for the methadone maintenance programme were potentially included; there were no exclusion criteria. During the period under study approximately 70 patients commenced maintenance therapy. Informed consent was sought and patients who agreed to participate were interviewed in the Drug Treatment Services clinic and were given $20 for their time spent in interview.

All clients who agreed to participate were interviewed by the same research assistant using a computerized version of the Composite International Diagnostic Interview (CIDI version 2.1) [13]. The CIDI inquires into symptomatology in the 12 months prior to interview and can furnish diagnoses (both current and pertaining to the 12 months prior to interview) using the International classification of diseases, 10th edition (ICD-10) and the Diagnostic and statistical manual, 4th edition (DSM-IV) criteria. This version of CIDI also inquires into symptoms of posttraumatic stress disorder (PTSD) and includes a Mini Mental State (MMS) examination [14] which earlier versions of CIDI did not. The CIDI has been widely used in epidemiological studies including the National Survey of Mental Health and Wellbeing [6] and has been shown to have excellent interrater reliability [15], [16].

The substance use module of the CIDI was not used as this would have significantly extended the length of the interviews (which were about 1 and 2 hours long without the substance use module). Also, the drug treatment workers took a thorough substance use history during the assessment for methadone maintenance therapy.

Most of our results are expressed as percentages. Where appropriate, comparison of observed rates was effected by χ² analysis, and comparison of continuous variables by t-test.

Results

In all, 44 males and 18 females were interviewed using the CIDI. The median age of the total sample was 24.6 years and the mean age was 27.4 years: the youngest person interviewed was 16.8 years and the oldest 55.1 years. Males were non-significantly older than females (28.5 years vs 24.7 years; t = 1.6, df = 60, p (two-tailed) = 0.11). All participants had a history of heroin use.

Participants had been on the methadone maintenance programme from less than 1 week (37.1%) to just under 6 months: for 62.9% of patients, this was their first time on a programme while 16.1% had been on the programme between 3 and 6 months.

Urine drug screens at the time of first assessment for the methadone programme showed that 42 of the 62 participants were poly-drug users: 35 showed evidence of cannabis use, 13 showed evidence of benzodiazepine use and two showed evidence of amphetamine use in the days prior to examination.

Only 3.2% of the cohort had no fixed abode at the time of being interviewed, 33.9% lived with spouse or partner, 24.2% lived with relatives, 24.3% lived with friends, and 16.2% lived alone or with their children. At the time of interview, 64.5% of the sample were unemployed.

The MMS, a screening instrument for cognitive impairment, is scored out of 30; scores of 23–24 are widely taken as the cutoff for probable impairment [17], pp.185–187]. The mean and median MMS scores for the sample were 27.3 and 28, respectively. Six patients showed some evidence of cognitive impairment: five patients had a MMS score of 22 or 23 while one individual (who had a diagnosis of hypomania) had a MMS score of 18. Only one patient was noted to be intoxicated at interview and this person did not fulfil criteria for a psychiatric diagnosis other than substance-use disorder.

Of the 62 patients, 15 (24.2%) did not fulfil ICD-10 criteria for a psychiatric diagnosis (other than substance abuse) at the time of interview or for the previous 12 months. Seven patients (11.3%) fulfilled criteria for only one diagnosis, 18 (29.0%) for two, five (8.1%) for three, 8 (12.9%) for four, and three patients (4.8%) each fulfilled criteria for five, six and seven or more diagnosis either at the time of interview or in the previous 12 months. Thus it may be seen that 75.8% of the sample fulfilled criteria for an ICD-10 diagnosis other than psychoactive substance-use disorder, while 64.5% fulfilled criteria for two or more diagnoses.

Eighty-nine per cent of female and 70% of male patients fulfilled ICD-10 criteria for a psychiatric diagnosis other than substance-use disorder. This difference is not statistically significant (χ² = 2.37, df = 1, p = 0.12).

Of the 62 patients interviewed, 53.2% fulfilled ICD-10 criteria for an affective disorder in the previous 12 months. Twenty-one had diagnoses of depressive disorder (20 had a moderate or severe depression and only one a diagnosis of mild depression), six had bipolar affective disorder, five fulfilled criteria for dysthymia as their only affective disorder diagnosis and one fulfilled criteria for hypomania. In total 67.7% of the sample fulfilled criteria for an anxiety disorder. Diagnoses included in the anxiety disorders group comprised generalized anxiety disorder (22 patients), posttraumatic stress disorder (19), panic disorder (12), obsessive–compulsive disorder (six), somatoform pain disorder (six) and a variety of phobias (22). It should be noted that many patients fulfilled criteria for more than one anxiety disorder. Twenty-eight, or 45.2%, fulfilled criteria for both an affective disorder and an anxiety disorder in the previous 12 months. It is of particular note that, at the time of interview, 19.4% of our sample fulfilled criteria for a current moderate or severe depressive disorder, and 46.8% a current anxiety disorder.

In addition to affective and anxiety disorders, one patient fulfilled criteria for a schizophrenic illness and was being treated for this. Two patients fulfilled criteria for an acute and transient psychotic disorder in the month prior to interview.

Treatment for psychiatric symptoms

Of the 47 patients who fulfilled criteria for an ICD-10 diagnosis other than substance-use disorder in the 12 months prior to interview, 15 reported that they had sought medical help for their psychiatric symptoms in that period. Of the 26 patients with a severe or moderate depressive disorder or bipolar disorder, 10 reported they had sought treatment as had two of the 14 patients who had an anxiety disorder alone. None of these patients had received treatment from the local public Area Mental Health Service in the 12 months prior to interview.

Comparison of the onsets of heroin use and psychiatric symptomatology

Patients who fulfilled diagnostic criteria for a psychiatric illness reported that they had first experienced psychiatric symptoms from less than 1 year to 32years prior to the interview, at a mean ageof 18.6 years. This group also reported that they had started using heroin from between 6 months and 30 years prior to interview, at a mean age of 21.7 years. Thus, for the group with a psychiatric diagnosis other than substance-use disorder, psychiatric symptomatology was reported to have preceded heroin use by just over 3 years on average. For 70% of those who fulfilled criteria for a psychiatric illness, psychiatric symptoms were reported to have preceded heroin use.

Fourteen patients with a psychiatric diagnosis other than substance abuse had anxiety diagnoses only, five had affective diagnoses only, and 28 had both. The mean ages of starting to use heroin of these three groups were 23.0, 23.9 and 20.8 years, respectively. In each group the onset of the first mental disorder preceded heroin use (nine of the anxiety only group, three of the affective only group and 21 of the combined group). On average, the onset of the first mental disorder preceded heroin use by 3.65 years in all cases with anxiety disorder diagnoses, and by 3.89 years in all cases with affective disorder diagnoses.

Discussion

In this study, the prevalence of psychiatric comorbidity in a methadone maintenance population in a semirural regional centre of Australia was investigated. The most significant finding was the high 12-month prevalence rates of anxiety disorder (67.7%), affective disorder (53.2%) and an anxiety and affective disorder together (45.2%). In Table 1 these findings are compared withthe 12-month prevalence rate of affective and anxiety disorders for the adult population (column 2) andthe 12-month prevalence of these disorder in that section of the population who also have a substance-use disorder (column 3) as found in the National Mental Health and Wellbeing study [6], which also used the CIDI. It can be seen that the prevalence of psychiatric disorder is seven to 10 times higher than that found in the adult population who did not fulfil criteria for a substance-use disorder and two to three times higher than the adult population who did (over 80% of whom had an alcohol-use or dependence disorder) [6]. Our results are consistent with the high rates of psychiatric symptoms found in previous studies [7], [9], [10]. The higher rates of symptomatology in female opiate users have also been foundpreviously [7], [9], [10]. Our finding that 19.4% of the sample were diagnosed with a moderate or severe depressive disorder and 46.8% with an anxiety disorder at the time of interview is also significant. Rounsaville [18] in a studyof 157 opiate addicts found that 17% were suffering from major depression at intake to treatment.

Table 1.

Prevalence of affective and anxiety disorders compared with the findings of the National Mental Health and Wellbeing study [6]

It is important to note that the sample in this study was not examined for DSM-IV axis II disorders as the focus was on the identification of more readily treatable disorders. It is recognized that the prevalence of personality disorders, particularly antisocial personality disorder, in the population who present for methadone treatment is higher than the general population. In one review the lifetime prevalence of antisocial personality disorder among opioid users and methadone patients rangedfrom 27% to 60% [19].

Why are the rates of psychiatric illness in this population so high? A distinction needs to be made between rates of psychiatric illness in those opioid users who do not seek treatment and those who do [20]. It has been noted previously that untreated opioid users have lower rates of depressive disorder, better family lives, fewer legal difficulties and less severe heroin use than those who seek treatment [20]. Galbaud du Fort et al. [21] observed that a comorbid major depression or anxiety disorder increased the probability of seeking treatment for many different disorders. It is also possible that those with psychiatric symptomatology are more likely to present for methadone treatment because they are more generally disorganized and experience greater difficulty maintaining a supply of heroin. An alternative explanation is that inconsistency of supply could lead to fluctuating symptoms of withdrawal and cause some of the psychiatric disorders seen on presentation or exacerbate existing psychiatric symptoms. A worrying possibility is that as a consequence of the recent reduction in the price of heroin [1], those with existing psychiatric illnesses may now be using it increasingly. This may lead to higher levels of psychiatric morbidity at presentation for methadone treatment in the future.

Two methodological limitations to this study, which may be related to one another, may have also contributed to the high prevalence rates found. First, patients in this study were between 1 week and 6 months into the methadone maintenance treatment when they were examined. It has been shown that those who are retained in a methadone maintenance programme experience a decrease in affective disturbance and anxiety over at least 18 months following commencing methadone treatment [22]. The second limitation concerns the quantity of methadone that patients were receiving at the time of examination. This ranged from 30 mg a day, usually early in treatment before stabilization of dosage has been achieved and when mood and anxiety symptoms can arise as a result of opioid withdrawal, to 100 mg a day. These mood and anxiety symptoms may be interpreted as disorders by the CIDI.

The finding that psychiatric symptomatology was reported to have preceded heroin use by over 3 years is of note and similar figures for affective and anxiety disorders been reported elsewhere [19]. A number of studies which have tried to disentangle the relationship between opioid dependence and psychiatric disorder suggest that the relationship may vary from one individual to another [19]. There are a number of possible interpretations of heroin use following the development of a psychiatric disorder. The two events may have no causal relationship at all. Early childhood trauma or disadvantage may lead to psychiatric ill health, including personality disorder, which in turn may predispose to heroin use. In addition, the use of drugs other than heroin may cause or exacerbate psychiatric illness.

We found that approximately one-third of those who were diagnosed as having a psychiatric illness reported that they had attended their general practitioner for their psychiatric symptoms in the 12 months prior to interview. These figures are higher than those found in a recent UK study where 20% of a cohort of 1075 adult dependent drug users had received treatment for a psychiatric disorder other than drug or alcohol dependence in the 2 years before intake to treatment [4].

How can the issue of the high prevalence of coexisting psychiatric illness in this population be addressed? There is some evidence that treatment of psychiatric symptoms can improve outcome for those embarking on methadone therapy [23]. The provision of weekly counselling has also been shown to produce better outcomes than the provision of methadone alone [24]. Because of the high prevalence of comorbidity and the difficulty in more than one service successfully engaging this population, many authorities suggest that at the level of specialist psychiatric and drug treatment services, each service must acquire skills to recognize and manage prevalent comorbid disorders among its patients [25].

The majority of patients with this comorbidity will be seen in the first instance by a general practitioner for assessment for methadone therapy. A comprehensive assessment by general practitioners commencing patients on methadone therapy should identify those whose psychiatric illness can be adequately treated by them with support from mental health services. It has also been suggested [18] that all patients should be screened after stabilization on methadone has been achieved, using a self-completed questionnaire such as the Beck Depression and Anxiety Inventories [25], [26]. Those who are suffering from a severe depressive disorder or psychotic disorder may require specialist treatment by mental health services.

As was mentioned in the introduction, this study was undertaken at the instigation of a group of general practitioners working in a regional setting. One approach would be for general practitioners working with this population to form a group (as has occurred in the Barwon region) so that education and support can be provided to them by drug treatment services and mental health services and unambiguous pathways to access to both these services can be made available. The option of shared care arrangements with mental health services is desirable, at least for a period of time until psychiatric symptoms are coming under control. Finally, it is important to acknowledge and rectify any remuneration issues, which would act as a disincentive to the general practitioner offering care for psychiatric illness as well as methadone.