The Influence of Client's Ethnicity on Psychotropic Medication Management in Community Mental Health Services

Abstract

Objectives: The aim of this study was to investigate whether people born in non-English-speaking countries differed from clients born in Australia on quality of medication management, measured by mean neuroleptic dose, method of administration, use of atypical antipsychotics and perceived compliance with medication, and to investigate the influence of matching the client with a case manager from the same ethnic background on these measures.

Method: Information about medication and perceived compliance was provided for 168 clients of five community mental health services in Melbourne. Chlorpromazine equivalent doses (CPZe) were calculated, and average dose, route of administration, percentage receiving atypical antipsychotics and perceived compliance with medication were analysed by country of birth and preferred language. Each analysis was repeated for clients matched to a case manager from the same ethno-linguistic background compared to those with ‘unmatched’ case managers.

Results: While non-English-speaking background (NESB) clients generally did not receive different dose sizes, those born in Vietnam had a lower mean dose. People born in a non-English-speaking country tended to be more likely than the Australian born to receive a depot injection, although this was not quite statistically significant. Twenty-seven percent of clients received an atypical neuroleptic; age was a significant factor, with older clients less likely to receive an atypical. There was no difference in receipt of atypicals or perceived compliance by country of birth, language or gender. Matching for a case manager of the same background had no effect except for route of administration, with matched clients less likely to receive depot medication than unmatched.

Conclusions: Generally, the ethnic background of clients had little influence on the quality of medication management they received from community mental health services.

The provision of psychotropic medication is a central aspect of community mental health service delivery. The management of medication type, dose size and compliance are important parts of service quality [1], but there has been comparatively little research investigating the influence of clients' ethnicity on medication management practices in mental health services. This is surprising given that research has shown that ethnic groups vary in their tolerance of psychotropic medications [2]. In particular, antipsychotic medications (also known as ‘neuro-leptics’) may have quite different pharmacokinetic and pharmacodynamic profiles in some ethnic groups. For example, research has indicated that Asian clients require lower doses for control of psychotic symptoms and, at doses commonly used in Caucasians, Asian clients may experience toxic side effects, such as extrapyramidal movement disorders (EPS) [2,3,4,5], which are a major cause of concern in all groups receiving antipsychotic medication.

There is some reason to expect that there may be ethnic differences in medication management practices. Language differences or communication difficulties may cause symptoms to be over- or under-diagnosed or misinterpreted, or side effects from medication to be missed [6,7]. Because of greater uncertainty about compliance, symptomatology and side effects, some clinicians feel that clients from ethnic minority backgrounds may be more likely to be treated conservatively (i.e. higher doses and more frequent use of depot antipsychotic preparations). It is also possible that due to these problems, clinicians might be reticent to place clients of non-English-speaking background (NESB) on the novel (‘atypical’) antipsychotics such as clozapine, risperidone and olanzapine. In addition, it could be expected that compliance with medication would be less among NESB clients due to greater difficulties in communicating about side effects and monitoring compliance [7] or differing explanatory models of illness [8].

A common recommendation for improving the cultural sensitivity of health services is the employment of staff from the same ethnic or language groups as clients [9,10]. A recent review of research concluded that, other things being equal, clients preferred a counsellor from the same ethnic background [11]. Some evidence exists that NESB clients who are matched to a counsellor from the same ethnic background or language group are less likely to drop out from services and have more contacts with service providers [12,13,14]. It could be expected that, if cultural differences or communication barriers lead to differences in medication prescribing for ethnic minority clients, ethnic matching of clients with primary case managers may lessen these differences.

Aim

This study examined antipsychotic medication use for a sample of clients from five community mental health services in Melbourne. We aimed to investigate whether people born in NESB countries, or who spoke a language other than English, differed from clients born in Australia on mean neuroleptic dose, method of administration, use of atypicals and perceived compliance with medication. We also aimed to investigate the impact on these variables of whether or not clients were matched with a case manager from the same ethnic background.

Method

Sample

As part of a broader evaluation project, medication information was collected from community teams at the South-West, Mid-West, Inner-North and North-West Area Mental Health Services in Melbourne in October 1997. Case managers were asked to provide details of the three main prescribed psychotropic medications, dose and frequency for clients. We aimed to include as many clients as possible at each service who were born overseas and spoke Croatian, Greek, Italian, Macedonian, Turkish or Vietnamese as their first language. We then selected a random sample of Australian-born clients from the same services (these were chosen by case managers from their own caseloads) to match with the NESB population.

Measures

Case managers rated the level of community functioning for each client using the Multnomah Community Ability Scale (MCAS). The MCAS is a 17-item scale developed by clinicians, managers and researchers in Oregon USA, with the intention of measuring outcomes of mental health services. It has been shown to have good reliability and predictive validity [15,16]. One item of the MCAS measures case managers' perceptions of client's compliance with prescribed medication on a five-point scale. Demographic data, including date of birth, country of birth, preferred language, and gender was obtained from the services' computerised databases. Case managers also provided an estimate of the client's fluency in English, and whether they spoke a language other than English (LOTE).

In order to contrast various doses of medications, oral and depot antipsychotic doses were converted to equivalent average daily doses of chlorpromazine (CPZe). Equivalences were calculated from a range suggested in the literature [17,18,19,20] and from clinical experience (TJRL). For the analysis of dose size, we excluded one case with an average dose of 4000 mg/day CPZe as an outlier, as the next highest dose was 1500 mg/day.

Statistical analysis

Differences on sociodemographic variables were investigated using t-tests. For route of administration, the proportion of NESB-born clients receiving a depot injection was compared with the proportion of Australian-born clients using the Chi-squared test. Mean medication dose for each group was compared using analysis of covariance (F-test). As the data for dose sizes were positively skewed, we used natural logarithm of dose size to conduct analyses. Proportions of each group receiving atypical antipsychotics were compared intially using the Chi-squared test. Further analysis was conducted by choosing a subset of clients matched for age and medication compliance, and using logistic regression to compare proportions receiving an atypical antipsychotic. Compliance with medication was compared using analysis of covariance.

For the analyses by country of birth, the sample was split into those born in Australia (54) or the United Kingdom (four clients) and those born in a non-English-speaking country (110). We further divided the NESB group according to whether they were born in a European country or an Asian country (Vietnam was the only Asian country included) where this distinction was relevant. For analyses by language, we split the sample into those speaking English, and those who preferred to speak a language other than English, and where relevant, this latter group was divided into those speaking a European language or Vietnamese. As it became evident that country of birth was confounded with age (those born in a non-English-speaking country were older), a separate analysis was conducted for a subset of the total sample. We matched each client born in Australia with a client born in a non-English-speaking country by age and medication compliance (based on the commonly used measure of Euclidean distance).

Results

Sociodemographic characteristics

Data were provided for 241 people. Seventy-one percent were diagnosed with a psychosis, 10.3% with bipolar disorder, 2.6% with a personality disorder, and diagnosis was missing for 9.5%. One hundred and sixty-eight (69.7% of the total) received neuroleptic medication, and the analyses in this study are restricted to this group. The major countries of birth were Australia (31.2%), Vietnam (20.7%), Italy (16.5%), Greece (11.4%), the Former Yugoslav republic of Macedonia (6.3%), Turkey (5.5%) and Croatia (3.4%). Preferred languages of clients were English (59.0%), Vietnamese (17.9%), Italian (7.3%), Greek (6.4%), Turkish (5.1%), Macedonian (1.7%) and Croatian (1.2%).

The mean age of the group was 40.7 (SD = 12.0), and 56% were male. There was no difference in the gender breakdown by birthplace or preferred language, but there was a significant difference for age. The overseas born group were older (mean age 43.3) than the Australian born (mean age = 35.5, t₁₅₇ = 4.06, p < 0.001), and those speaking a European language (mean age = 46.4) were older than English speakers (mean age = 38.4) and Vietnamese-speakers (mean age = 34.4, F_2,160 = 13.0, p < 0.001). For the 110 clients born in a non-English-speaking country, 35 (31.8%) saw a case manager from the same ethnolinguistic background and 75 (68.1%) did not.

Route of administration

Table 1 shows route of administration by birthplace and preferred language. Of the sample, 39.9% received a depot only and 55.3% received an oral neuroleptic only. Interestingly, only eight people (4.8%) received both oral and depot medication, and seven of these were born in a non-English-speaking country. There was no difference in the proportion receiving a depot (with or without an oral) by age, birthplace or preferred language, although there was a trend for the overseas born to be more likely to receive a depot (χ² ₁ = 3.06, p = 0.08).

Table 1.

Route of administration by birthplace and preferred language

We then analysed likelihood of receiving a depot by whether the client saw a case manager from the same ethno-linguistic background or not. Clients matched to a case manager of the same background were less likely to be on a depot (χ² ₁ = 7.85, p = 0.005).

Average daily antipsychotic dose

The mean dose was calculated for each birthplace and language group, and is shown in Table 2 by administration route. There was a significant relationship between total dose and both age (r = −0.23, p = 0.003) and gender (t₁₃₆ = 2.59, p = 0.01), with younger clients and males having higher doses. As it might be expected that clients of Asian backgrounds would receive smaller doses, we first analysed doses for those born in Vietnam compared with the rest of the sample, after controlling for age and gender. This analysis showed that the Vietnamese-born received a significantly lower total dose (Vietnamese mean dose 244 mg/day, all others mean dose 342 mg/day, F_1,156 = 2.5, p = 0.03).

Table 2.

Mean dose by birthplace, preferred language and administration route

That Asian people require smaller doses of psychotropic medication is presumably due to genetic differences among regional population groups, rather than ethnic identity, cultural difference or communication barriers. In order to explore possible cultural differences in dose, we excluded Asian-born clients and then analysed average dose by birthplace and preferred language. After controlling for age and gender, there was no difference in oral, depot or total dose size by country of birth or preferred language. There was also no effect when considering whether clients were matched to a case manager of the same ethnic background.

Atypical antipsychotics

Table 3 shows the percentage of each birthplace and language group receiving an atypical antipsychotic (clozapine, olanzapine or risperidone). Forty-five (27.1%) were on a novel drug (risperidone 14.5%, clozapine 10.2% and olanzapine 2.4%) with the rest (72.9%) on a traditional (‘typical’) neuroleptic.

Table 3.

Percent receiving atypical antipsychotics by birthplace and preferred language

The proportion of English-speakers receiving an atypical medication was no different to the proportion of those speaking a LOTE (χ² ₁ = 0.29, p = 0.59). However, there was a trend for a greater proportion of those born in Australia to receive atypical medications when compared to those born in a NESB country (χ² ₁ = 3.68, p = 0.055). Those born overseas were less likely to be on a novel drug. There was also a strong effect for age, with those receiving an atypical medication being significantly younger (t₁₅₉ = 2.18, p = 0.03). However, age and birthplace were confounded, with NESB people being older than the Australian-born group.

In order to investigate this relationship further, we matched clients for age and medication compliance so that all 54 clients born in Australia were matched with the same number born in a non-English-speaking country. Logistic regression was conducted with sex, age, and medication compliance as covariates. This analysis showed a significant effect for age (p < 0.01), but no effect for birthplace, gender or compliance rating. Clients matched with a case manager of the same ethnic background were no more likely to receive an atypical drug than NESB clients with unmatched case managers.

Perceived compliance with medication

Case managers rated the compliance of clients with medication on a five-point scale which comprised one of the 17 items of the Multnomah Community Ability Scale. Higher scores indicated greater perceived compliance. Mean compliance ratings by birthplace and preferred language and administration type are shown in Table 4.

Table 4.

Mean rating of perceived compliance (on a five-point scale) by administration route, birthplace and preferred language

There was no difference in perceived compliance with medication either by birthplace, preferred language, or gender. There was a small correlation of age with compliance (r = 0.19, p = 0.017), greater compliance being associated with older clients.

The difference in compliance between clients on depots and those on oral medications was not significant, despite a numerical trend for those on depots to be seen as less compliant (t₁₆₆ = 1.88, p = 0.06). For both the Australian-born and English-speakers, those on depots were rated as less compliant than those who were not. In contrast, there was no difference in compliance between oral and depot delivery routes either for clients born in a NES country or those speaking a language other than English. Clients who were matched with a case manger of the same ethnic background received compliance ratings no different to those with ‘unmatched’ case managers.

Conclusions

We aimed to investigate the impact of ethnicity, as determined by country of birth and preferred language, on the management of antipsychotic prescribing in a community mental health setting.

In terms of route of administration, there was a trend for the NESB born to be more likely than the Australian born to receive a depot medication. If this trend reflects a true relationship (a bigger sample might be required to confirm this), then it supports the notion that clinicians may err on the side of caution with respect to route of administration when faced with cultural, ethnic or communication barriers. This is supported by the finding that clients matched to a case manager from the same ethnic background were much less likely to receive a depot than NESB clients with an ‘unmatched’ case manager.

When considering average daily doses of antipsychotics, younger clients and males received higher doses. Clients born in Vietnam received significantly lower doses (mean dose = 244 mg/day) than other clients (mean dose = 342 mg/day). This finding matches those of studies of other Asian groups [4]. After excluding Asian-born clients, no difference was found in total dose for those born in a non-English-speaking country compared to the Australian born. Overall, there was no difference in dose size when considering ethnic matching to a case manager. Dose sizes for both Asian born clients and those from other backgrounds appear consistent with guidelines for good clinical practice, with the mean dose for non-Asians at the lower end of the recommended range for maintenance doses of 300–600 CPZe mg/day [5,21].

We have hypothesised that clients of NESB might have restricted access to the atypical antispychotics. The findings revealed, however, that there was no difference in likelihood of being prescribed an atypical antispychotic by country of birth, preferred language, or ethnic matching with case manager. Older clients were less likely to be on an atypical antispychotic, but this was a trend independent of ethnicity and gender. This suggests that access to newer agents in this community clinic setting is equitable with regard to ethnicity.

We found no difference in perceived compliance by country of birth, preferred language, sex, or ethnic matching with case manager, but younger clients were perceived to be less compliant. For clients born in Australia or who preferred to speak English, those receiving depot medication were rated as less compliant. This finding is consistent with clinical guidelines for depot medication to be used where compliance with orally administration is uncertain. However, there was no similar relationship between compliance and depot medication for the NESB-born or those speaking a language other than English.

It is possible that depots tend to be used more for NESB clients because there is greater uncertainty about compliance or occurrence of symptoms for NESB people due to cultural differences or communication difficulties. The fact that clients matched to case managers from the same ethnic background did not receive depots more frequently supports this argument. However, if the concern was about undetected symptoms, one would also expect to find higher average doses as for NESB people, which was not the case.

There were some limitations of this study. First, the data collection pro-forma included space for only three medication types (although some case managers added more); some clients may have received more than three different medication but these would generally not have been recorded. The small proportion of clients recorded as receiving both oral and depot medication suggests that multiple medication types were underestimated. Second, the information about medication was provided by case managers based on information recorded in each client's file notes. Case managers less familiar with medication types and doses may have recorded these inaccurately. It would have been useful to cross-check the information recorded by case managers with that recorded in files, but this was not possible due to resource constraints. Finally, there may be some question about how representative the sample was. While we tried to include all clients from the targeted ethnic backgrounds, we were unable to get information for all. In addition, although case managers were asked to choose Australian-born clients at random, we do not know how random this process was. Future research will need to consider these issues.

The results of this study show that ethnicity does not appear to be related to the quality of medication management in the community mental health services included. Whether this finding is generalisable to other services is open to speculation; the services included in this study have a long history of working with clients from ethnic minority backgrounds and are generally attuned to issues of cultural sensitivity in service delivery. In addition, a significant number of staff included in this study were bilingual/bicultural and they may have been more attuned to cultural issues than staff generally. Similar studies in a wider range of services are necessary in order to make judgements about the quality of medication management for NESB clients in the mental health service system as a whole.