Review of the forensic psychiatry service at Zomba Mental Hospital and a comparison of forensic psychiatry services in Malawi and Scotland

Abstract

This retrospective case-note study describes the demographic details, offences and diagnostic characteristics of 283 patients admitted to Zomba Mental Hospital, Malawi, after early diversion from the Criminal Justice System between May 1997 and February 2007. Given the historical links between Malawi and Scotland and the ongoing involvement of Scottish psychiatrists in the development of psychiatry in Malawi, a comparison is drawn between the provision of secure psychiatric treatment in these low- and high-income countries. Consistent with Scottish prison mental health team referrals and Scottish high secure psychiatric patients, the Malawian patients were predominantly men, poorly educated and drug misusing, but alcohol was less often a problem. Affective disorders were rarely diagnosed, as is also the case in the equivalent Scottish populations. In Malawi, there appears to be a problem with aftercare because many of the patients had been detained before. The length of detention was very short, other than an exceptional finding for substance misuse – a mental disorder with no Mental Health Act equivalence in the UK. Absconding was the commonest outcome. Malawian patients diverted from the Criminal Justice System are treated on general adult psychiatry wards where few secure measures can be offered. Further, there is little capacity for follow-up, no community compulsion legislation exists and psychiatric expertise is limited to a few disparate sites. However, the service is developing and the data in this study represent a small part of that ongoing programme.

Introduction

Malawi, in south central Africa, is bordered to the east and south by Mozambique (with which it shares Lake Malawi) and by Tanzania and Zambia to the north and west. Malawi has a population of 12.7 million but a life-expectancy of only 41 years.¹ Human immunodeficiency virus (HIV) infection is endemic, with 15–17% of pregnant women being HIV+.² Malawi is a poor country too; the average annual income per capita is US $160³ and half the population live below the poverty line.

The prevalence of schizophrenia, bipolar disorder and postnatal depression in sub-Saharan Africa is similar to that in high-income countries,^4,5 but Malawi, like its neighbours, struggles to provide a mental health service. Recruiting and retaining staff is difficult, services are over-centralized and drug supply inconsistent.⁶

There is a general hospital in all but four of Malawi's 28 administrative districts but only three psychiatric units. Of these, only Zomba Mental Hospital in the south has a psychiatrist. The other psychiatric facilities, like much of Malawi's health service, are staffed by nurses and paramedical clinical officers.⁷

Mental illness in Malawi used to be managed in Zomba Central Prison until Zomba Mental Hospital was built in 1953. The hospital now has 330 beds and admits between 1500 and 2000 patients per year. It has a catchment area of eight million and is staffed by Malawi's two psychiatrists with five clinical officers and 20 nurses.

Malawi has a lengthy association with Scotland that started with Dr Livingstone in 1859 (Dr Livingstone was born in Blantyre, Lanarkshire, after which the capital of Malawi's southern region is named.) Livingstone was inspired to follow ‘the smoke of a thousand villages to where no missionary had ever been’ but led British colonial commerce into central Africa. The Nyasaland protectorate was then established and only secured independence, so becoming Malawi, in 1964. Ngwazi Dr Hastings Kamuzu Banda assumed the life presidency having previously studied medicine at the University of Edinburgh and worked as a general practitioner in Renfrew. Further, as a result of Malawi's two psychiatrists having trained in both Malawi and Scotland, the development of psychiatry in Malawi is supported by the Scotland Malawi Mental Health Education Project (see www.smmhep.org.uk).

The demographic, offence and mental disorder characteristics of the forensic patients treated in Zomba Mental Hospital are described here. Further, because of the association between Malawi and Scotland referred to above, these data are compared with the analogous facilities of The State Hospital, Carstairs, Lanarkshire (which provides care and treatment for approximately 240 patients under conditions of special security) and the mental health service at HMP Perth (which serves as a prison for medium-term sentenced prisoners as well as a remand prison for Tayside and Fife).

Method

The Malawian Mental Treatment Act 1948 includes four orders, of which two can act as hospital diversions from the Criminal Justice System. The police can petition for a 30-day order (along with a medical officer or relative, there being no mental health officer equivalent in Malawi) and the Court can impose a Reception Order (which can be indefinite and has no formal need for review). Patients treated under both these Mental Treatment Act Orders were included in the present study.

Between May 1997 and February 2007, 283 such patients were detained in Zomba Mental Hospital. Records relating to their management were identified and data analogous to that in previous published literature were extracted. The diagnosis was that recorded in the medical notes. When reviewing the data, the strength of association between categorical variables was verified using the χ ² test on SPSS version 14 (2006, SPSS Inc, Chicago, USA). These data were compared with the 1997 State Hospital Survey⁸ and the 2003/4 review of HMP Perth mental health team referrals.⁹

Results

Sociodemographic details

The average age of the patients at Zomba Mental Hospital was 30.4 years (range 15–65 years); 91.5% were men and only 22.6% had been educated beyond primary school level (Table 1). Of the patients, 19.8% had misused alcohol and 23% had misused illicit substances. Of those who misused illicit substances, 98.3% reported using cannabis (‘chamba’) with only 1.7% reporting ‘another substance’ (this is likely to be khat, an amphetamine-like stimulant derived from the Catha edulis tree). Deliberate self-harm (DSH) was not a feature of any of the admissions.

Table 1

Index offence and educational standard (χ ² 20.77, df 9, P = 0.014)

	Murder	Assault	Damage to property	Other	Total (%)
No schooling	36	16	16	31	99 (35.0%)
Primary school	40	22	17	41	120 (42.4%)
Secondary school	5	13	17	27	62 (21.9%)
University	0	1	0	1	2 (0.7%)
Total	81	52	50	100	283

Clinical details

Because there are only three psychiatric institutions in Malawi (none of which can currently offer specialist forensic facilities), there are no intrahospital transfers. The patients were all admitted following early hospital diversion from police custody or court. The index offences are categorized in Table 2.

Table 2

Index offence

Index offence	%
Murder	28.6
Assault including sexual	18.4
Damage to property including arson	17.7
Other including breach of the peace	35.3

Schizophrenia was the most common diagnosis (35.5%), but a near equivalent number of patients (32.5%) were detained because of ‘substance misuse’ (The Mental Health [Care and Treatment] (Scotland) Act 2003 does not allow for detention in cases of substance misuse alone). No patients were diagnosed with personality disorder and multiple diagnoses were not recorded (Table 3).

Table 3

Diagnosis

Diagnosis	%
Schizophrenia	35.5
Depression	3
Mania	0
Learning disability	0
Personality disorder	0
Organic syndrome	0
Substance misuse	32.5
None	8.5
Epilepsy	8.1
Stress disorder	1.8

A significant proportion of the patients had been detained before: 31.8% under a Temporary Treatment Order (a civil detention order which can last up to 12 months, applied for by a relative with a recommendation from a medical officer) and 4.6% under a Reception Order. Of those admitted because of substance misuse, 44.6% had been detained before (Table 4).

Table 4

Diagnosis and previous formal admission (χ² 22.76, df 12, P = 0.030)

	No previous admission	Previous admission under TTO	Previous admission under RO	Total
Not mentally ill	22	2	0	24
Schizophrenia	61	32	7	100
Substance abuse	51	37	4	92
Depression	2	1	0	3
Epilepsy	16	5	2	23
Stress disorder	1	4	0	5
Other	27	9	0	36
Total	180	90	13	283

The majority of patients (75%) were admitted for 1–6 months (Table 5) and the length of admission was not affected by diagnosis. On discharge, unless they had absconded, the patients were returned to the Criminal Justice System. Not infrequently, patients were admitted for a number of weeks before it was concluded that they were ‘not mentally ill’.

Table 5

Length of admission

Length of admission	%
Less than 1 day	1.4
Up to 1 week	12.7
Up to 1 month	40.3
Up to 6 months	35
More than 6 months	10.6

Schizophrenia was associated with violence against the person and damage to property. There was also an apparent association of epilepsy with murder. No patients were diagnosed with mania. Diagnoses and index offences are correlated in Table 6.

Table 6

Diagnosis and index offence (χ ² 36.37, df 18, P = 0.006)

	Murder	Assault	Damage to property	Other	Total
Not mentally ill	10	2	3	9	24
Schizophrenia	30	23	11	36	100
Substance abuse	16	16	28	32	92
Depression	1	1	0	1	3
Epilepsy	13	5	3	2	23
Stress disorder	1	1	0	3	5
Other	10	4	5	17	36
Total	81	52	50	100	283

Regardless of the offence committed, forensic admissions are treated on the general adult psychiatry wards. Only routine security measures are available, staff numbers are comparatively low (particularly at night) and there is no capacity to nurse individuals in isolation. Perhaps unsurprisingly then, the majority of patients absconded; 45.6% were discharged versus 54.4% absconded. Index offences and absconding rates are summarized in Table 7.

Table 7

Index offence and absconding rates (χ² 0.67, df 3, P = 0.881)

	Discharged	Absconded	Total
Murder	39	42	81
Assault	25	27	52
Damage to property	22	28	50
Other	43	57	100
Total	129	154	283

The admissions, although formal, were usually short. Of those detained following a murder, 42% were in hospital for less than one month. Index offences and duration of subsequent admission are correlated in Table 8.

Table 8

Index offence and length of admission (χ² 35.04 df 12 P = 0.000)

	Less than one day	Up to one week	Up to one month	Up to six months	More than six months	Total
Murder	3	6	25	28	19	81
Assault including sexual	0	4	21	22	5	52
Damage to property including arson	0	7	27	14	2	50
Other including breach of peace	1	19	41	35	4	100
Total	4	36	114	99	30	283

Discussion

This study is limited in a number of ways. The institutions compared have differing remits, diagnosis maybe inaccurate in Malawi because of the service limitations mentioned and the prominence of some conditions could be an artefact of only single diagnoses being recorded.

The average patient age was 30.4 years, which is lower than the mean age of 40 years found in a representative British general adult psychiatry inpatient study.¹⁰ However, the average age of Malawian forensic psychiatric inpatients does compare to the commonest age range referred to the HMP Perth mental health team⁹ and the average age of patients in the State Hospital, Carstairs.⁸ Further, 91.5% of the Malawian sample was men so this population appears analogous to that treated in secure institutions in Scotland. However, the life-expectancy in Malawi is similar to the average age of a psychiatric inpatient in Britain so the results will be skewed.

Only primary schooling from six to 14 years of age is compulsory in Malawi. Thereafter, students must be selected and pass an examination if they are to attend secondary school. As a result, educational opportunities are limited but, despite this, there is still an inverse correlation between educational achievement and most crimes, particularly those of violence (Table 1).

Of the patients, 19.8% had misused alcohol and 23% had misused drugs. Of the drug misusers though, only 1.7% had taken something other than cannabis. In the State Hospital, 48.5% had a history of heavy alcohol use and 46.9% had used illegal drugs (9.5% intravenously). In HMP Perth, 69.6% had a history of problematic substance misuse. Alcohol and substance misuse were considerably lower and less diverse in the Malawian sample than in the analogous Scottish populations.

It is of note that the rates of substance misuse and schizophrenia in this sample, when taken together, are very similar to the overall rates of schizophrenia in the State Hospital. Duration of admission and rates of previous admission are similar for these two conditions in Zomba Mental Hospital as well. Perhaps then, a proportion of the admissions ascribed to substance misuse are actually schizophrenic presentations. This anomaly is, of course, not unique to Malawi. It is not clear whether those admitted with substance misuse had psychotic symptoms.

The rates of DSH contrast greatly between Malawi and Scotland. While 61.4% of State Hospital patients and 42.6% of the HMP Perth sample had deliberately harmed themselves (varying from scratches of the wrists to life-threatening injury), DSH was not a feature of any of the Malawian presentations. Suicide is illegal in Malawi and DSH does not appear to be a culturally sanctioned expression of distress. This is demonstrated by the preponderance of ligature points and slatted glass windows at Zomba Mental Hospital which are not used by patients to self-harm.

Our own survey of Zomba Mental Hospital patients with intact reality testing revealed that about half believed witchcraft was responsible for their mental illness (this is not delusional but in keeping with the culture¹¹) and half supposed they had a brain disorder. None thought that adverse childhood experience was causal.

In the State Hospital, 5.4% of patients had a primary diagnosis of personality disorder (26.2% of patients in English Special Hospitals have such a diagnosis, probably because of differences in the use and interpretation of Mental Health Act Legislation in England/Wales and Scotland) and 47% of the HMP Perth sample had a diagnosis of neurosis/personality disorder, but none of the Zomba Mental Hospital sample was so diagnosed. The Malawian Mental Health Act does not formally define mental disorder other than to specify that this includes mental defects (learning disability). Accordingly, responsibility for the management of personality disordered individuals usually remains with the Criminal Justice System. Very occasionally, Zomba Central Prison has been asked to take over detention of patients who were thought too dangerous to be treated on the hospital wards.

Maintaining patients out of hospital is very difficult in Malawi because of the service limitations mentioned. As a result, and despite the centralization of psychiatric services, 31.8% of the patients had been admitted previously to Zomba Mental Hospital under a Temporary Treatment Order and 4.6% under a Reception Order. In the State Hospital, 22.4% of patients had had at least one previous State Hospital admission (formal admissions elsewhere were not recorded in the quoted survey). In Zomba Mental Hospital then, the rate of re-admission under detention appears to exceed that in the State Hospital and all the Zomba Mental Hospital re-admissions followed an offence being committed in the community. Only 48.9% of State Hospital admissions actually followed an offence; 35.3% were transfers from another hospital.

The short duration of admission might also result in comparatively rapid re-admission as most of the patients were only admitted for between one and six months (in the State Hospital, the average length of admission is five years with a range of six months to 30.5 years). Finally, because of the degree of family cohesion in Malawi and the involvement of many Malawians in subsistence agriculture, many people can stay at home until they are quite unwell. As a result, many Malawian psychiatric inpatients have comparatively severe conditions. Patients who did not abscond were discharged back to the Criminal Justice System but the Psychiatric Service was only made aware of the outcome of three such cases.

The Malawian data suggest that there is an association between diagnosis and index offence: schizophrenia is associated with violent offending on a much larger scale than the other diagnoses, including substance misuse. Such an association was also reported in a study¹² that demonstrated a high risk of violence among men with schizophrenia: 9% of non-fatal personal assaults, 21% of offences of damage to property, 30% of the arson cases and 11% of the murders in this remand prison sample having been committed by men with schizophrenia despite the general population prevalence of schizophrenia being only 0.1–0.4%. However, other published studies suggest that the contribution of mental illnesses to community levels of serious violence in the community is small. For example, one comprehensive American survey¹³ demonstrated that schizophrenia or major affective disorder increased the risk of violence to 8% from a base rate of 2%, and the MacArthur Violence Risk Assessment Study¹⁴ suggested that discharged psychiatric patients are not more dangerous that the general population. A further study¹⁵ reported that screening positive for psychosis in a cross-sectional survey of persons in households conveyed an attributable risk of violence of only 1.2%, with hazardous drinking and antisocial personality disorder conveying the most risk (50.9% and 24%, respectively).

In a study¹⁶ that re-visited the MacArthur Violence Risk Assessment, it was concluded that patient monitoring and the avoidance of substance misuse are the critical factors in reducing the risk of re-offending in discharged psychiatric patients (as a result, to ensure treatment adherence in the community, the authors advocate the use of ‘leverage’ measures like tying welfare benefits to treatment participation). Options like these are not available in Malawi, so the service cannot influence concordance in discharged patients to modify risk or re-admission rates.

In this study, there also appears to be a particular association between epilepsy and murder but the validity of this diagnosis maybe questionable. Although detention was by virtue of epilepsy, the case-notes did not indicate that the violence took place as a result of peri-ictal phenomena. For example, the offence behaviour was often organized and the patients could recall the event.

Conclusion

There are similarities between the forensic psychiatric patients in Scotland and Malawi. Both groups are predominantly men, have poor scholastic histories and misuse drugs. Affective disorders are rarely diagnosed. However, there are a number of differences because, in Malawi, there are more admissions of previously detained people suggesting an aftercare issue; the length of detention is very short (other than an exceptional finding for substance misuse) and alcohol misuse is less frequent. Detention for substance misuse alone is common in Malawi, although this is not allowed under the Scottish Mental Health Act. In Malawi, the data suggest an association between schizophrenia and violence and epilepsy and murder, but the latter diagnosis may not always have been accurate. Finally, the absconding rate in Malawi, particularly for murderers, greatly exceeds that in Scotland. A number of these differences illustrate the need for continued service development in Malawi.

Footnotes

Acknowledgements

The authors are grateful to Drs P Bennie, R Stewart and H Johnsson of the Scotland Malawi Mental Health Education Project for their suggestions and critical reviews.

References

World Health Organisation. Mortality Country Factsheet, 2006. See www.who.int (last checked 20 January 2010)

United Nations. AIDS epidemic update, 2007. See www.unaids.org (last checked 20 January 2010)

The World Bank. Country Brief, 2008. See web.worldbank.org (last checked 20 January 2010)

Odejide

, Oyewunmi

, Ohaeri

. Psychiatry in Africa: an overview. Am J Psychiatry 1989;146:708–16

Stewart

. Maternal depression and infant growth: a review of recent evidence. Mater Child Nutr 2007;3:94–107

Kauye

. Management of mental health services in Malawi. Int Psychiatry 2008;5:29–31

Kauye

, Mafuta

. Country profile: Malawi. Int Psychiatry 2006;4:9–11

Thompson

, Bogue

, Humphreys

, Owens

, Johnstone

. The State Hospital Survey; a description of psychiatric patients in conditions of special security in Scotland. J Forensic Psychiatr 1997;8:263–84

Friel

, White

. Outcome of referral to a prison mental health team. Med Sci Law 2006;46:321–7

10.

Simpson

, Middleton

. Fast stream psychiatric rehabilitation after recent hospital closure. Psychiatric Bull 1994;18:613–4

11.

Crumlish

. Insight, psychopathology and global functioning in schizophrenia in urban Malawi. Br J Psychiatry 2007;191:262–3. See Podcast www.rcpsych.ac.uk/download/podcast/crumlish_mix.mp3 (last checked 20 January 2010)

12.

Taylor

, Gunn

. Violence and psychosis. I – risk of violence among psychotic men. BMJ 1984;288:1945–9

13.

Swanson

, Holzer

, Ganju

, Jono

. Violence and psychiatric disorder in the community: evidence from the Epidemiologic Catchment Area surveys. Hosp Community Psychiatr 1990;41:761–70

14.

Steadman

, Mulvey

, Monahan

, Violence by people discharged from acute psychiatric inpatient facilities and by others in the same neighborhoods. Arch Gen Psychiatr 1998;55:393–401

15.

Coid

, Yang

, Roberts

, Violence and psychiatric morbidity in the national household population of Britain: public health implications. Br J Psychiatr 2006;189:12–19

16.

Torrey

, Stanley

, Monahan

, Steadman

. The MacArthur Violence Risk Assessment Study revisited: two views ten years after its initial publication. Psychiatric Ser 2008;59:147–52