Social Psychiatry and Psychiatric Epidemiology

, Volume 43, Issue 5, pp 387–391

No evidence for restrictive care practices in Māori admitted to a New Zealand psychiatric inpatient unit: do specialist cultural teams have a role?


  • Shailesh Kumar
    • Dept. of Psychiatry, Kingsley Mortimer UnitNorth Shore Hospital
    • Dept. of Psychiatry, Kingsley Mortimer UnitNorth Shore Hospital
  • Alexander Simpson
    • Dept. of Psychiatry, Kingsley Mortimer UnitNorth Shore Hospital
  • Jesse Fischer
    • Dept. of Psychiatry, Kingsley Mortimer UnitNorth Shore Hospital
  • Elizabeth Robinson
    • Dept. of Psychiatry, Kingsley Mortimer UnitNorth Shore Hospital

DOI: 10.1007/s00127-008-0320-6

Cite this article as:
Kumar, S., Ng, B., Simpson, A. et al. Soc Psychiat Epidemiol (2008) 43: 387. doi:10.1007/s00127-008-0320-6



To ascertain the presence, and describe the pattern and extent, of restrictive care practices in the treatment of mental health inpatients in a rural New Zealand unit.


Retrospective data was anonymously extracted from patient records at Rotorua Hospital (Rotorua, New Zealand). Data sets were compiled from 300 consecutive patient admissions between January 2000 and December 2001. The demographic and diagnostic characteristics extracted were gender, age, ethnicity (Māori or non-Māori classification only), primary diagnosis, length of hospital stay, seclusion, medication on discharge, dosage of antipsychotic medication if given, referral to psychotherapy, voluntary/involuntary status on admission, and readmission rates.


After controlling for other clinical variables, ethnicity was not associated with specific diagnoses, increased use of seclusion, and involuntary status on admission or higher readmission rates. Māori patients were more likely to receive antipsychotic medication and at higher doses than non-Māori. Māori were less likely to be referred to psychotherapy services and had shorter lengths of stay.


There was no evidence of widespread restrictive care practices against Māori, although the disparities in antipsychotic prescription and psychotherapy referral suggest some restrictive care practices do exist. The use of specialist cultural teams in general mental health services may prevent restrictive care practices.

Key words

ethnic minority groupsrestrictive care practicetranscultural psychiatryMāori

1 Introduction

Improving the care of indigenous peoples with mental illness is a vital challenge for mental health services. This includes New Zealand, where Māori, the indigenous people, make up 12–15% of the population [24]. Factors that may contribute to an increase in psychiatric morbidity in indigenous peoples include destruction of social infrastructure, changes in lifestyle, rapid urbanization, poverty, cultural alienation and loss of identity, family dislocation, traumatisation and increased alcohol and drug use [3, 28].

While these factors contribute to our understanding of higher psychiatric morbidity within ethnic minorities, concerns are also being expressed that such groups may present later and receive poorer or more restrictive care. This includes reduced access to services, compulsory treatment under legislation [7, 18], increased use of certain types of medication at higher dosages [2, 11, 21], under prescribing of appropriate medications [6, 17, 20, 29], poor utilisation of interventions such as ECT [4], and the excessive use of seclusion and restraint [5, 8, 23].

We have used the term restrictive care to describe psychiatric practices that reduce the autonomy and choice of the individual. Examples of restrictive care include the use of legislation for involuntary admissions, readmissions, the use of restraint and seclusion and higher doses of medication. Whilst the use of such practices should be minimised, they remain an important aspect of care that can restore competence and autonomy in people with severe mental illness. If used excessively, restrictive care may decrease patients’ voluntarism and impede their recovery [12], resulting in an underutilisation of available treatments, the inappropriate use of treatments or the excessive use of coercive practices. Patients’ perceptions of how they were treated by mental health clinicians and their legal status may be significant in their recovery. A Swedish study of short term outcomes of brief inpatient care found that patients who reported being treated well by staff during their hospital stay were more likely to show greater degrees of improvement in global assessment of functioning (GAF) scores [30]. Involuntary admitted patients reported higher levels of perceived coercion compared to those who are admitted voluntarily [14]. If restrictive care practices are applied more frequently to certain populations, such as indigenous peoples and ethnic minorities [19], it may suggest that services are not adequately meeting their needs.

Several studies have shown that Māori receive poorer or more restrictive care than non-Māori and mental health services are perceived as punitive rather than assisting in recovery [9, 10, 15, 22, 26, 27]. Most international and New Zealand studies have only considered individual aspects of psychiatric care. No study, to the best of our knowledge, has looked at ethnic differences across all aspects of psychiatric care. A major New Zealand study [10] of classification and mental illness outcomes commented on aspects of care that may be associated with restrictive care practices, such as seclusion and antipsychotic usage, but did not examine the issue of restrictive care practices in its own right.

The following retrospective study will examine the use of restrictive care practices in consecutive admissions on a general adult psychiatric unit in rural New Zealand. Our hypothesis is that Māori experience more restrictive practices as an inpatient than non-Māori.

2 Methods

The study was conducted at Rotorua Hospital in Rotorua, New Zealand. The catchment area population is 105,000, 35% of whom are Māori, and consists of two towns, Rotorua and Taupo, and the neighbouring rural area. The hospital has one acute adult inpatient psychiatric unit with 14 general beds and two seclusion beds. Patients are admitted to the unit on a voluntary and involuntary basis for any acute psychiatric condition for which community interventions have failed, with or without risk to self or others. Community interventions are offered through a Community Mental Health Team which operates with five residential homes and several respite admission beds as alternatives to hospitalisation. There are no child or forensic admissions. The main diagnostic system is DSM-IV based [1].

There is a bicultural approach on the inpatient unit with full time Māori nurses and Māori cultural workers (Whai Mānaaki), some of who are proficient in Te Reo Māori (Māori language). There is also access to a tohunga (an expert in Māori spiritual and health matters). Iwi (tribal) support for the research project was obtained. All Māori patients, at all points of contact with the Mental Health Services are given the option of having a Whai Mānaki present with them and most patients opt for it. Consultations invariably start with a Mihi (Māori welcome) followed by Karakia (Māori prayer) and finish with Karakia. At all points, Whai Mānaki ensured that Māori traditions and the cultural rights of Māori patients were adhered to and respected.

Retrospective data was extracted anonymously from patient files on 300 consecutive inpatient admissions between January 2000 and December 2001. The demographic and diagnostic characteristics extracted were gender, age, ethnicity (Māori or non-Māori classification only), primary diagnosis, length of hospital stay, seclusion, medication on discharge, dosage of antipsychotic medication if given, referral to psychotherapy, voluntary/involuntary status on admission, and readmission rates.

2.1 Statistical analyses

Frequencies, medians and means are presented to describe the demographics and diagnostic characteristics of the patients. T tests and chi-squared tests were used to compare the differences in age and gender between the two ethnic groups. Diagnoses were grouped into five categories: adjustment disorder, anxiety disorder, mood disorder, schizophrenia and others. A generalised logit regression model was used to investigate the differences in diagnoses between Māori and non-Māori when controlling for age and gender.

Multiple linear and logistic regression analyses were used to investigate whether there were differences in readmissions, the use of compulsory admissions, use of seclusion, doses of antipsychotic medication, rates of referral to psychological services and the total length of stay between Māori and non-Māori when other factors were controlled. Factors used in the models included age, gender, diagnosis, time between onset of illness and admission and where appropriate, the number of readmissions and compulsory status under the Mental Health Act on admission. SAS version 9.1 was used for all analyses (SAS Institute Inc., Cary, NC, USA).

3 Results

Details of the 300 patients, of whom 125 were Māori and 175 non-Māori, are shown on Table 1. One Māori patient received no diagnosis, and was excluded from subsequent analysis involving diagnosis. There was no significant difference between the proportions of males and females in each group (45.6% of Māori were female, 42.9% of non-Māori were females, chisq (1df) = 0.22, P = 0.6). Māori patients were younger with a mean age (SD) of 30.8 (10.87) years compared with 38.3 (13.74) in non-Māori (t289 = 5.11, P < 0.0001).
Table 1

Comparisons of diagnoses between Māori and Non-Māori


Schizophrenia (%)

Mood disorder (%)

Anxiety disorder (%)

Adjustment disorder (%)

Other diagnoses (%)

P value from regression model


62 (50.0)

38 (30.6)

10 (8.1)

9 (7.3)

5 (4.0)



64 (36.6)

80 (45.7)

2 (1.1)

15 (8.6)

14 (8.0)


Age ≤30

67 (54.0)

30 (24.2)

7 (5.7)

13 (10.5)

7 (5.7)


Age >30

59 (33.7)

88 (50.3)

5 (2.9)

11 (6.3)

12 (6.9)



79 (47.3)

63 (37.7)

3 (1.8)

11 (6.6)

11 (6.6)



47 (35.6)

55 (41.7)

9 (1.8)

13 (9.9)

8 (6.1)


aOne Māori patient had no diagnosis, so was excluded from analysis

After controlling for age and gender there was a higher prevalence of schizophrenia in Māori, but this did not reach statistical significance (P = 0.06). Younger patients on the other hand were more likely to be diagnosed with schizophrenia and less likely to have a mood disorder (P = 0.0005) (Table 1). No differences were found in the readmission rates for either ethnic group (P = 0.15), gender (P = 0.5) or age (P = 0.27) after controlling for the other demographics (Table 1).

A multiple regression model was used to investigate the factors associated with the length of admission. A log transformation was performed so that assumptions of normality were not violated. Non-Māori stayed longer than Māori (P = 0.01), with a median of 8 days compared to a median of 5 days for Māori (Table 2). There was a trend towards length of admission (median no. of days) being associated with diagnosis: mood disorders (9), other disorders (7), schizophrenia (6), adjustment disorders (5) and anxiety disorders (4) (P = 0.06).
Table 2

Results of comparison of Māori and non-Māori from regression models


Māori (n = 125)

Non-Māori (n = 175)

Results from regression models

Median length of stay (range)

5 (0–38)

8 (0–71)

Β = −0.46SE = 0.17P = 0.01

Number of patients placed in seclusion

38 (31%)

33 (19%)

OR = 1.61

95% CI (0.86–3.01)

P = 0.1

Antipsychotic dose on discharge (CPZ equivalent)

  0 mg

25 (20.0%)

55 (31.4%)

OR = 1.90

95% CI (1.11–3.25)

P = 0.02

  0–200 mg

31 (24.8%)

53 (30.3%)


  >200 mg

69 (55.2%)

67 (38.3%)


Compulsory status on admittance

61 (49%)

70 (40%)

OR = 0.84

95% CI (0.50–1.41)

P = 0.5

Referral for psychotherapy

11 (9%)

63 (36%)

OR = 0.16

95% CI (0.07–0.35)

P < .0001

Number of patients readmitted within 12 weeks of previous discharge

64 (51%)

90 (52%)

Β = 1.35

SE = 0.93

P = 0.15

A logistic regression model with age, sex, diagnosis, number of readmissions and time (days) between onset of illness episode and admission included as explanatory variables found that being placed in seclusion was associated with the number of readmissions [OR = 1.07 (1.03–1.12) P = 0.0009]. Those with more readmissions were more likely to be placed in seclusion. Ethnicity [OR = 1.61 (0.86–3.01) P = 0.1] and the other variables were not found to be associated with use of seclusion.

The average chlorpromazine equivalent antipsychotic dose was categorised into three levels: 80 (27%) did not receive such medication, 84 (28%) received up to 200 mg of CPZ equivalent antipsychotic dose and 36 (45%) more than 200 mg CPZ equivalent antipsychotic dose. An ordinal logistic regression was used to investigate which factors were associated with the dose. Māori were more likely to be given antipsychotics, with 100 (80%) patients receiving it compared to 120 (69%) in non-Māori [OR = 1.90 (1.11–3.25), P = 0.02], with a higher median (range) of 311.0 mg (50–1,500) compared with 249.4 mg (3–1,500) CPZ equivalent antipsychotic dose. Age and gender had no effect on whether or not a patient received antipsychotics, or on the dosage they received. Ethnicity had no effect on the type of antipsychotic patients received (atypical antipsychotics, typical antipsychotics, or a combination of the two), (P = 0.7). There was also no association found between ethnicity and the administration route (oral vs. depot with or without oral antipsychotic). 88.5% of patients admitted under the MHA were given antipsychotics, compared to 61.5% of voluntary patients. They were also more likely to be on a higher dose, with 70.2% on more than 200 mg chlorpromazine equivalents compared with only 26.0% of voluntary patients (P = 0.0001).

A multiple logistic regression model including sex, age, ethnicity, diagnosis and number of readmissions as variables, found that ethnicity was not associated with compulsory status on admission [OR = 0.84 (0.50–1.41), P = 0.5]. Diagnosis (P = 0.006) and the number of readmissions [OR = 0.94 (0.90–0.98) P = 0.003] were both associated with the use of legislation for compulsory treatment in the community.

A multiple logistic regression was used to investigate which factors were associated with referral for psychotherapy in the community. Females (42, 32%) were more likely to be referred to psychological services than males (33, 20%) [OR = 2.13 (1.15–3.98)]. Māori were less likely to be referred than non-Māori [OR = 0.16 (0.07–0.35)]. Age was not a significant variable. The probability of referral decreased with the number of readmissions [OR = 0.86 (0.76–0.97)]. Those with adjustment disorders [OR = 6.13 (2.09–17.92)] and anxiety disorders [OR = 5.88 (1.02–33.8)] were more likely to be referred for psychotherapy than patients with schizophrenia.

Power analysis: although the study had over 90% power to detect differences between Māori and non-Māori in the proportions in seclusion and those admitted under compulsory status of 20%, the differences found were smaller and therefore the power to detect these was between 40 and 60%.

4 Discussion

Restrictive care practices were studied in a rural New Zealand inpatient psychiatric unit. In 300 consecutive admissions, Māori ethnicity did not have a significant association with length of hospital stay, the use of seclusion, compulsory status on admission or higher readmission rates, after controlling for other clinical variables. Māori inpatients were more likely to be prescribed antipsychotics and at higher doses than non-Māori inpatients but there was no difference in the class or administration route of antipsychotic medication. Māori inpatients were less likely to be referred for community psychotherapy services, which is consistent with a previous report [15].

The findings that Māori inpatients were younger and more likely to have a diagnosis of schizophrenia are consistent with the recent Mental Health classification and outcomes study (CAOS) that found psychotic disorders were the commonest diagnoses received for Māori (63%) compared to European (39%) in an inpatient setting [10]. In this study, there was a non-significant trend for non-Māori inpatients to have longer lengths of admission than Māori, but longer inpatients times were also associated with mood disorders. Despite longer admission lengths, non-Māori readmission rates were similar to Māori. This may indicate issues with access at the point of receipt of care [25] or better social support networks operating for Māori minimizing readmission rates [16], a finding not described previously in New Zealand. The lack of difference in compulsory status on admission contrasts with a previous study [9]. Our finding that Māori comprised 35% of the local population but 42% of hospital admissions suggests that Māori are disproportionately over represented in the inpatient population, which is consistent with the national trend in New Zealand [10].

Māori received higher doses of antipsychotic medication, a finding seen in other ethnic minority populations overseas [2, 11, 21]. Possible explanations include biological variability, the treatment of more severe or later presenting illnesses, prejudices operating against Māori or a restrictive care practice. The last factor may be unlikely as there was no difference in depot antipsychotic use. Depot medication is frequently seen as a restrictive care practice, given its intrusion into civil liberties, and is reported to be used more frequently in other ethnic minority groups [2, 18]. On the other hand we found that the higher usage of antipsychotics in Māori remained significant even when diagnosis was controlled implying that the usage of antipsychotics for Māori with “non-psychotic diagnoses” was more liberal than non-Māori. This finding may suggest that treatment approaches adopted by mainstream mental health services may be fundamentally different for Māori compared to non-Māori populations.

The study does have limitations. Its retrospective nature did not allow the collection of extra data. However a prospective design may have changed clinical practice. The variables analysed for restrictive care were derived from the literature and only have face validity. Other more subtle variables may have been missed. Restricted care practices were only studied in a general adult inpatient setting and did not extend to the region’s corresponding community mental health centres or other specialist units, which might have practices that may impact on inpatient treatment. This may limit the generalisability of our findings.

Under representation in community care [10], increased utilization of inpatient treatment by crisis teams [27] and poorer treatment during and after incarceration [22] may still be important factors. We note that the confidence interval for odds ratio for hours spent in seclusion was wide [OR = 1.61 95% CI (0.86–3.01)] possibly because the numbers in seclusion were small. The study also does not offer any explanation for the trend towards a disparity in the diagnosis of schizophrenia between Māori and non-Māori. Finally, whilst this study might be able to highlight possible disparities in psychiatric care, it cannot offer explanations for such disparities or apparent lack of disparities.

One hypothesis for the near absence of an ethnicity effect on restrictive practices may be the close integration of Māori Mental Health Teams with the main stream Mental Health services. We speculate that a strong and active Māori mental health team that works alongside non-Māori clinicians as would have been evident from the description of care provision above may have been instrumental in minimizing restrictive care practices. Since all New Zealand public psychiatric services have Māori Mental Health Teams, we were unable to compare our findings with a region that did not have them.

This study also highlights the need for improving staff attitudes and practice towards patients of ethnic minorities in psychiatric services. Indeed, this change at the “grass roots level” may be more important than “top down” legislative changes in reducing restrictive practices, such as seclusion and restraint [13].

5 Conclusion

The present study of a general inpatient unit in rural New Zealand did not find differences in restrictive practices between Māori and non-Māori. Such practices have been previously reported in the treatment of Māori patients [10, 26, 27]. Whether or not this is seen in other services, such as community mental health centres, and other regions warrants further research. The positive impact of Māori Mental Health Teams may not yet be fully realised. Antipsychotic medication use and dosage, poor access to psychotherapy for Māori patients and other, more subtle restrictive care practices may need further prospective study.

Copyright information

© Steinkopff Verlag Darmstadt 2008