Abstract
It has been found from mental health data that patterns of diagnosis, trends over time and use of psychiatric services vary among different ethnic groups within New Zealand [1–3]. Although little systematic research into Maori mental health has been carried out, available data suggest that rates of admission to hospital for any reason are higher for Maori than non-Maori [1]. Regional differences seem to exist, with the Ministry for Maori Development (Te Puni Kokiri) reporting higher rates of Maori admissions to southern regional psychiatric hospitals compared with other areas of New Zealand [2]. These results are difficult to interpret, however, given that the Ministry for Maori Development report also found that when Maori became unwell, they were not necessarily treated within their home region [2].
First admission rates to psychiatric services for Maori have been reported as remaining static between 1984 and 1993 [2,3]. However, the validity of the data on which these claims were made has been called into question making it difficult to comment definitively on these findings. Compared with non-Maori, however, readmission rates for those of Maori descent have continued to rise over the same time period [2]. Patients of Maori descent are more likely than non-Maori to be referred for psychiatric admission by welfare, the law or relatives, with non-Maori being more likely to be referred by psychiatrists, doctors or outpatient services. These findings suggest Maori and non-Maori have different pathways into New Zealand's psychiatric services.
High rates of health problems other than mental health, have been found among people with psychiatric illness when compared with the general population [4]. Many of these health problems are also associated with lower socioeconomic status per se, in which the Maori population is overrepresented.
Suicide risk for people with psychiatric illness is higher than that of the general population [5,6]. For example, 44% of a recent Dunedin sample of bipolar disorder patients reported having attempted suicide at least once [7]. Traditionally, Maori suicide rates have been lower than non-Maori, but more recent studies have found little difference between the two ethnic groups suggesting a possible rise in suicide rates among Maori [5,8].
Published data on trends in Maori health are based on several assumptions: (i) that the data gathered is accurate; (ii) that Maori concepts of health and wellbeing are encompassed within the Western medical paradigm, which is the most common model of assessment; and (iii) that a standardised definition of ethnicity has been used. In many studies these assumptions are not correct, making it difficult to compare findings and leading to widespread underreporting of Maori morbidity and mortality data [9,10].
Recognition by the Ministry for Maori Development of the lack of accurate Maori mental health research has led to this area being identified as a health gain priority area [2,3]. The study outlined here was developed in response to that call. We describe the preliminary results of a project examining Maori and non-Maori first admissions to psychiatric units within Otago over a 2-year period, 1990–1992. The paper outlines the demographic data and preliminary descriptive findings on Maori and non-Maori admissions.
Method
A retrospective, case review study of all first admissions to psychiatric units and wards within Healthcare Otago's catchment area (Cherry Farm Hospital, Dunedin Public Hospital and Wakari Hospital) between January 1990 and December 1991 was carried out. First admission was defined as the first inpatient stay in any one of Healthcare Otago's units. As such, the data described here will include some patients who had previous admissions to other parts of New Zealand or overseas. First admissions were identified from the hospital admission records, which record all admissions within the service. Admission records for the first 8 months of 1990 were not available for one Dunedin hospital unit.
Sample
The records of 259 psychiatric admissions were identified. Patients ranged in age from 15 to 45 years. The records of individuals who were inpatients at the time of review were excluded. Admissions were divided into Maori and non-Maori, with Maori ethnicity defined within this study as being any individual identified, either by themselves, as part of the admission procedure or at any point within the body of the notes, as being of Maori descent. The Southern Regional Ethics Committee approved this study.
Review process
The case notes were reviewed by the principal author (LE). Descriptive information concerning the index admission was gathered including demographic data, ethnicity, length of stay, diagnoses, treatments, comorbidity and sources of referral. Data were also collected on the above areas for all subsequent admissions. Subsequent readmission data will not be presented here. Positive family history was defined as having one or more family member(s) identified within the clinical notes as having a psychiatric illness.
Data collection
Collected data were entered (by LE) onto coding sheets with all identifying features eliminated. Where ambiguities existed, in terms of diagnoses given and clinical data recorded within the notes, details were checked with the supervising psychiatrist (AESW). Collated data were then entered onto an SPSS database spreadsheet (SPSS Inc., Chicago, IL, USA).
Statistical analysis
Age standardisation to the 1991 New Zealand census data for sole Maori (referring to those people of Maori descent who identified Maori as being the only ethnic group to which they belong) and non-Maori population were carried out using the direct method. Chi-squared tests were used to test for differences between Maori and non-Maori groups. Odds ratios (OR) and 95% confidence intervals (CI) are presented for some comparisons. Fisher's exact tests were used where the expected value was less than 5. Poisson regression was used to compare the suicide rate in Maori and non-Maori. Analyses were carried out using the SPSS for Windows Statistics Program (SPSS Inc., Chicago, IL, USA).
Results
Sample
First admissions between January 1990 and December 1991 for 259 patients, aged between 15 and 45 years, were identified.
Using New Zealand sole Maori and non-Maori population figures, the age standardised rates of Otago's admissions per 1000 population were 4.0 for Maori (95% CI = 1.7–11.5) and 1.0 for non-Maori (95% CI = 0.4–1.2).
The sample included 42 (16.2%) Maori and 217 (83.8%) non-Maori. Of the Maori sample 36 (86%) were male compared with 125 (58%) of the non-Maori sample. This difference was significant (χ2 = 11.82, df = 1, p < 0.0001). Age at first admission was not significantly different between Maori (mean age = 27.2 years, SD = 7.6) and non-Maori (mean age = 28.1 years, SD = 7.5).
Details of marital and occupational demographics are shown in Table 1. The differences were not significant for marital status (p = 0.051) or for occupation (p = 0.11). The OR for being on a benefit was 1.86 (95% CI = 0.9–3.8) for Maori compared with non-Maori.
Marital status and occupations of first psychiatric admissions (January 1990 to December 1991)
In terms of educational attainment, 30 (71.4%) of all Maori first admissions did not report any formal educational qualifications compared with 92 (47.9%) of non-Maori. Of those patients with tertiary qualifications (polytechnic or university study), six (14.3%) were Maori and 49 (22.6%) non-Maori (χ2 = 11.9, df = 1, p < 0.001).
Outlines of referral source patterns and contacts with psychiatric services prior to admission are presented in Table 2. Significant differences were found between Maori and non-Maori with regard to referral sources (Fisher's exact, p = 0.02) and prior contact with psychiatric services (χ2 = 0.14, df = 1, p = 0.01). For Maori 22 (52.4%) were referred by legal services compared with 52 (24%) for non-Maori. Maori were also more likely to have had a prior admission outside of the Otago region with 21 (50%) Maori having had a prior admission compared with 52 (24%) for non-Maori.
Referral sources and prior psychiatric service contact amongst first psychiatric admissions (January 1990 to December 1991)
No significant difference was found between Maori and non-Maori with regard to admission status (χ2 = 3.24, df = 1, p = 0.07). Of those Maori who had never had a previous admission outside of Otago, 13 (62%) had admissions of informal status. Informal admission status was considered to be admissions that were voluntary and not under the Mental Health (Compulsory Assessment and Treatment) Act 1992 [11]. In contrast, of those Maori who reported admission(s) outside of the Otago region four (19%) had informal status admissions (χ2 = 0.65, df = 1, p = 0.42). For non-Maori, of those with no prior admissions, 103 (70.5%) had informal admission status compared with 26 (37.1%) of those with a previous admission outside the Otago area (χ2 = 2.39, df = 1, p = 0.12).
Diagnoses for Maori and non-Maori first admissions are summarised in Table 3. A significant difference was found between Maori and non-Maori with regard to diagnosis (Fisher's exact, p = 0.43). At the time of discharge eight (19%) Maori had no diagnosis or their diagnosis was undecided compared with 21 (9.7%) of non-Maori.
Diagnosis at discharge for first psychiatric admissions (January 1990 to December 1991)
Of all identified first admissions, 113 (43.6%) reported a positive family history for psychiatric illness, with similar rates found for Maori (n = 21, 50%) and non-Maori (n = 42, 42%). In addition, no significant difference in the total number of family members identified as having a psychiatric illness was found between Maori and non-Maori families (Maori: mode = 1, range = 1–4; non-Maori: mode = 1, range = 1–8).
Health problems, other than psychiatric, did not differ significantly between the groups with 19 (45.2%) Maori and 71 (32.7%) non-Maori reporting problems (χ2 = 1.24, df = 1, p = 0.26).
Suicide attempts were recorded for 24 (57.1%) Maori patients compared with 89 (41.0%) non-Maori patients, giving a relative risk (RR) for suicide attempts of 2.1 for Maori versus non-Maori (95% CI = 1.6–2.8). Lethality or intent of suicide attempts was not recorded. Both Maori and non-Maori males, however, made more suicide attempts than females with the a age group 15–25 years being the time of greatest risk. No significant difference was found between Maori and non-Maori with reference to the number of suicide attempts (mean number of attempts: Maori = 2.6; non-Maori = 1.8).
Discussion
To our knowledge this paper reports the first systematic case review study examining Maori first psychiatric admissions within the Otago region. Of particular note was the finding that Maori rates for a first psychiatric admission were four times that of non-Maori. Maori male admissions were also overrepresented in our sample, a finding in keeping with other studies [1–3]. The reason why Maori are overrepresented among first psychiatric admissions in Otago is unclear. Such high rates may reflect the poorer overall health status seen among Maori and their overrepresentation in lower socioeconomic groups. Both factors are known to be associated with higher rates of mental illness.
In contrast, only six Maori females were admitted for inpatient care. The reasons for this are equally unclear and the small sample size makes it difficult to comment on this finding with any confidence. Such numbers are, however, in marked contrast to the high rates of self-reported depressive symptomatology found in a community sample of Maori women [12]. This discrepancy, between high rates of depressive sympatomatology in the community and low inpatient admissions for depressive illness, raises the possibility of poor identification of psychiatric illness in Maori women by health professionals and poor access to and/or unwillingness by Maori women to accept psychiatric hospitalisation as it currently exists.
Research carried out in other parts of New Zealand has suggested that diagnoses of schizophrenia and/or psychotic illnesses are the most common diagnoses given to Maori at the time of discharge, with one study reporting the rate of diagnoses of depressive disorders among Maori first admissions at one-half the rate of schizophrenia [2,9]. In keeping with research into other minority groups, our findings however, suggest affective disorder rates among Maori are higher than rates of psychoses or schizophrenia, paralleling the rates of psychiatric disorders seen among people of European descent [10,13]. Indeed depressive disorder was the most common diagnosis given to both Maori and non-Maori first admission cohorts in our study. Maori were also more likely in this study to be given no diagnosis or to have their diagnosis deferred at the time of discharge. Such findings suggest a delay in reaching a definitive diagnosis and thus instigating treatment for Maori.
Why diagnostic patterns for Maori in Otago should differ to those of the rest of the country is unclear, but the differences raise several interesting possibilities. First, it is possible that this local finding represents a national trend, reflecting an overall improvement in the standard of diagnostic expertise among New Zealand psychiatrists for Maori. It would be interesting to examine diagnoses given to current psychiatric admissions to see if a difference in diagnostic patterns is occurring. Alternatively, it is possible that Otago's Maori psychiatric population mix differs from that of the rest of New Zealand, having a higher than expected proportion of families with affective illness and/or lower rate of psychosis. No data exist to suggest this latter explanation is true, although a smaller proportion of the Otago population is Maori compared with the rest of New Zealand. Maori in Otago do not differ from Maori in other parts of New Zealand with regards to demographic factors. Given that over one-half of the Maori sample had a prior admission to another inpatient unit outside of the Otago region this may influence the diagnostic patterns seen here. The diagnostic patterns seen here warrant further investigation.
In terms of family psychiatric illness and marital status, Maori and non-Maori inpatients did not differ from each other. Similar total numbers of affected family members were identified for both groups, and both cohorts were more likely to be single (59.5%) than married. While identification of psychiatric illness in a family member was usually made on the basis of another family member's report, experience by the authors in other studies suggest families are very accurate in identifying family members who are ill [7]. Findings on marital status are similarly in keeping with other studies, which have found an association between currently being single, and the presence of psychiatric illness [14].
Maori were overrepresented in the lower socioeconomic groups, with significantly more Maori receiving a government benefit compared with non-Maori. Maori reported higher rates of having no qualifications and lower rates of tertiary qualifications compared with non-Maori, in keeping with previous findings [15]. While the results most likely reflect the overall disadvantaged position Maori hold within New Zealand society, they may also reflect the socio-economic, educational and occupational disadvantage associated with psychiatric illness [4,13]. Such reasons may also hold true in terms of explaining the high rates of other (non-psychiatric) health problems found among Maori within this sample.
In terms of admission status, Maori were more likely to have had an earlier admission to a psychiatric inpatient service outside of the Otago region compared with non-Maori. This may reflect a greater mobility among Maori compared with non-Maori. The fact that the age of first admission to Otago hospitals did not differ between Maori and non-Maori is also of interest, as it raises the possibility of an earlier age of onset of psychiatric illness for Maori, given their higher proportion of earlier admissions outside of Otago. Maori were also more likely to be admitted under a formal admissions status (non-voluntary) and referred by the law, compared with non-Maori. This finding supports earlier research and may reflect poor access and/or unwillingness of Maori to accept mental health care resulting in delayed presentation [2,3,8]. These findings should, however, be tempered by the knowledge that Dunedin runs a regional forensic service, accepting admissions from a wide catchment area. Furthermore, Maori are overrepresented within New Zealand's penal system and are more likely to access mental health services via forensic services. Our findings may well reflect this gateway to psychiatric services for Maori men.
Reported suicide attempts were common in both Maori and non-Maori cohorts, especially in males aged 15–25 years with Maori having a relative risk of suicide attempts double that of non-Maori. While the suicide risk for psychiatric patients is well documented as being higher than that of the general population, Maori suicide rates have traditionally been lower than non-Maori [5] [6] [7,16]. Our finding that Maori report more suicide attempts than their non-Maori counterparts contradicts earlier studies and suggests a disquieting trend. While a distinction must be drawn between suicide attempts and rates of completed suicide, it is probable that our finding reflects an underlying rising suicide rate among Maori psychiatric patients as there is no evidence to suggest a falling rate for non-Maori. Other sources also support a rising suicide rate in Maori, especially Maori youth aged 15–25 years [8].
In designing this study, an emphasis was placed on high quality data collection and accuracy. Nevertheless, a number of methodological limitations must be acknowledged. Data were gathered retrospectively from case-note review. Thus, the authors were reliant on the accuracy with which the information was gathered at the time of admission.
Ethnicity is defined by self-determination and from any recorded reference to ethnicity in the hospital notes. While such a definition might be considered broad, it represents a deliberate attempt on the part of the authors to maximise the number of Maori patients correctly identified as such. Definitions of ethnicity have varied over time [17] with various definitions being used depending on trends. The type of definition of ethnicity used is important as it will influence the denominator used in such things as age and population standardisation. Varying ethnicity definitions influences the size of a particular ethnic group and so the prevalence of illness within this group. Identification of ethnicity by self-determination is in keeping with current trends within New Zealand, such as the collection of census data. Nevertheless, our ethnicity data should be read with the caveat that definitions of ethnicity used between 1990 and 1992 may not mirror those used in the current study, and that the original data collected may not have been accurate. It is not possible to know what definitions of ethnicity were used by medical professionals at the time of admission, they may well have differed from the broad definition used by the authors. Given the well-documented underreporting of Maori admissions to date, however, it is also true that our use of a broad definition of ethnicity is likely to increase the accuracy of data [9]. On balance, the authors believe that any bias in the findings reported here for Maori would still be in the direction of under- rather than an overestimation.
The data set was incomplete as admissions from one hospital unit for the first 8 months of 1990 were missing. The authors believe, howevere that the vast majority of first admissions to Otago psychiatric inpatient services were captured within the reported data due to the high standards of record keeping.
Additionally, data gathered were based on a Western biomedical paradigm and may not have included aspects considered by Maori to be important for Maori health and wellbeing. The use of a biomedical paradigm was useful in enabling the gathering of clear descriptive information that could later be developed into further research by those working within the Western medical and Maori paradigms of health and wellbeing. Finally, funding constraints precluded any independent reliability checks being conducted on the case-note reviews carried out by the principal author (LE).
In summary, we report here preliminary demographic and descriptive data on a cohort of Maori and non-Maori first admissions to Otago psychiatric wards between December 1990 and January 1992. The preliminary findings highlight higher than expected rates of Maori admissions to psychiatric services and diagnostic patterns for Maori paralleling non-Maori with depressive disorders being most commonly seen. Maori were more likely to have no diagnosis or an undecided diagnosis at the time of discharge. Maori reported higher rates of welfare support and higher rates of having no qualifications. The Maori cohort was also likely to have other health problems and was more likely to have attempted suicide with a higher relative risk. The authors hope to publish more details from this study including information on diagnosis and comorbidity and subsequent admissions in the near future.
Footnotes
Acknowledgements
The Health Research Council of New Zealand, Te Oranga Tonu Tanga, Ngai Tahu Health Research Unit and the Affective Disorders Research Group, the University of Otago supported this research.