First-episode psychosis: An epidemiological survey comparing psychotic depression with schizophrenia

2.1. Background 

The PACE (population-adjusted clinical epidemiology) process had been developed and refined over a 20-year period in haematological malignancy (Proctor and Taylor, 2000, Proctor, 2002, Charlton et al., 1997). PACE uses a robust registry system to record comprehensive prospective data on the presentation, management, and outcomes of all cases of a disease entity in a defined locality. In 1998 the haematology methodology was adapted as a practical system of continuous assessment and evaluation for all patients with a first-episode psychosis in Northumberland. PACE was developed for the purposes of service evaluation, audit, and with the overall aim of improving services (Proctor et al., 2004).

2.2. Subjects 

All patients aged 16 years and older seen by Northumberland consultant psychiatrists with a diagnosis of first-episode psychosis, according to ICD-10 criteria were included. There was no upper age limit. Patients with physical health problems and/or co-existing substance misuse were also included, unlike most RCTs which normally exclude such patients.

The only exclusion criteria were: presentation outside of the Trust area with a first-episode psychosis, past history of treated psychosis, or a diagnosis of dementia at presentation (those who had mild cognitive impairment, but whose main clinical problem was psychosis were included).

2.3. Procedure 

From October 1998 all patients with a first-episode psychosis were identified by the consultant psychiatrists in Northumberland, and referred to the PACE team. A basic data form, completed either by consultants or PACE staff, gathered information on demographics, risk assessment and ICD-10 diagnosis. PACE staff gathered data from inpatient and outpatient secondary care medical notes. The data were entered onto an Access database. All patients' secondary care medical notes were searched annually on the anniversary of first presentation. Information on demographics, contact status, hospital admissions, risk assessment, medication and treatment were collected annually. PACE staff had no contact with patients, and were not involved in their care or management.

Monthly reminders were sent to all consultant psychiatrists, asking for names of patients, (both inpatients and outpatients), who had recently presented with a first-episode psychosis. PACE staff also screened the local computerised patient information system for potential patients who had been discharged from hospital. All adult and old age consultant psychiatrists were involved in identifying potential patients.

2.4. Analysis 

Each patient was given a clinical ICD-10 diagnosis by their consultant psychiatrist when they first presented to secondary care services. This diagnosis was used to place patients into different ICD-10 diagnostic groups, and the outcome data for these groups were analysed using Microsoft Access 2000 and SPSS 14.0 for Windows. Continuous data were analysed using the independent samples t-test, and the ordinal data using the chi-square test.

2.5. Confidentiality 

Confidentiality was strictly maintained. All patients were given unique PACE identification numbers. No individual patient was ever identified as data were always presented in the form of group outcomes.

3.1. Incidence and demographics 

Over the 7-year period from October 1998 to October 2005 there were 540 cases of ICD-10 first-episode psychotic illness (309 males and 231 females). Significantly more patients presented with psychotic depression than schizophrenia (p<0.05, 95% CI=1.09, 1.89, RR=1.44). There were 105 cases of first-episode psychotic depression (F32.3), which was 19% of all first-episode psychosis. There were 73 cases of first-episode schizophrenia (F20), which was 13% of all first-episode psychosis. Table 1 shows the number of cases of each disorder by gender and age. Psychotic depression was statistically less likely to occur in patients under 36 years (p<0.01, 95% CI=1.41, 3.04, RR=2.07). Schizophrenia was significantly more likely to occur in patients under 36 years (p<0.05, 95% CI=1.01, 2.40, RR=1.56). No significant gender differences were found overall in the psychotic depression group (p=0.25, 95% CI=0.80, 1.59, RR=1.13) or the schizophrenia group (p=0.125, 95% CI=2.26, 0.91, RR=1.44). However, a relationship was observed between age and gender. Males under 36 years were significantly less likely to have a diagnosis of psychotic depression (p<0.01), whereas, males over 64 years were significantly more likely to receive this diagnosis (p<0.01). Females between the ages of 36 and 64 years were more likely to be diagnosed with psychotic depression, than females in the other age groups (p<0.01).

Table 1. Incidence at presentation of different diagnostic groups in first-episode psychosis

		No. of cases	Male	Female
	Psychotic depression	105	48	57
	Aged <36	31	15	16
	Aged 36–64	43	18	25
	Aged >64	31	15	16
	Schizophrenia	73	48	25
	Aged <36	42	37	5
	Aged 36–64	19	9	10
	Aged >64	12	2	10
	All first-episode psychosis	540	309	231
	Aged <36	251	176	75
	Aged 36–64	160	85	75
	Aged >64	129	48	81

In the psychotic depression group, 72% (n=76) of patients were still in contact at year 1. For the schizophrenia group, this was 76% (n=118) at year 1. There were 10 deaths in the psychotic depression group and 7 deaths in the schizophrenia group.

3.2. Hospital admissions 

An independent samples t-test showed no significant differences between the two groups for number of hospital admissions (p=0.76) and total number of hospital admission days (p=0.78). No significant differences were found for the number of days spent in hospital per year after presentation (p=0.37, p=0.45, and p=0.18 for years 1, 2, and 3 respectively). Table 2 shows that 81 patients with psychotic depression were admitted with a total of 10,025 days (mean 95 days), and 47 schizophrenia patients were admitted to hospital with a total of 10,844 days (mean 149 days). Chi-square test revealed that the Mental Health Act was used significantly more times with the schizophrenia group than the psychotic depression group (p<0.05).

Table 2. Sum of hospital days by diagnosis, gender and age (between October 1998 and October 2006, n=355 patients admitted)

	ICD-10	All first-episode psychosis			Aged <36			Aged 36–64			Aged >64
		Total	Males	Females	Total	Males	Females	Total	Males	Females	Total	Males	Females
	Psychotic depression
	No. admitted	81	37	39	21	12	9	33	13	20	27	12	15
	No. of days	10025	4896	4128	2794	1292	1502	3579	1933	1646	3652	1671	1981
	Mean days	124	132	106	140	108	167	108	149	82	135	139	132
	Schizophrenia
	No. admitted	47	34	13	27	23	4	14	9	5	6	2	4
	No. of days	10844	9163	1681	7656	7362	294	2278	1494	784	910	307	603
	Mean days	231	270	129	284	320	74	163	166	157	152	154	151
	All first-episode psychosis
	No. admitted	355	203	152	155	110	45	116	59	57	84	34	50
	No. of days	54725	32161	22564	30043	20989	9054	12133	6216	5917	12549	4956	7593
	Mean days	154	158	148	194	191	201	105	105	104	149	146	152

3.3. ECT 

Of the psychotic depression group, a total of 18 patients received ECT (mean age of 63). There were no gender differences. All were reported to respond well to ECT at the time.

3.4. Medication 

Of the psychotic depression group 92% (n=97) were treated at some point with antipsychotic medication. 46% (n=45) were treated with at least 2 different antipsychotics, 60% (n=58) of patients were prescribed first-line antipsychotic medication prior to or at presentation, 95% (n=92) patients received an antipsychotic within the first 3 months. 70 patients received atypical antipsychotic medication. 98% (n=103) were treated with antidepressant medication. 53% (n=56) patients were treated with at least 2 different antidepressants, 77% (n=81) were prescribed a first-line antidepressant prior to or at presentation, 17% (n=18) within the first 3 months.

Of the schizophrenia group, 100% (n=73) received an antipsychotic, and 45% (n=23) were prescribed an antidepressant. None received ECT. 85% (n=62) were prescribed atypical antipsychotic medication.

3.5. Physical health problems 

Significantly more patients in the psychotic depression group had physical health problems at presentation (56%; n=59; p<0.01). This difference was still significant at year one follow-up (p<0.05). 28% (n=21) of the schizophrenia group had physical health problems. In both groups, the most common problems were cardiovascular, followed by neurological and respiratory disorders.

3.6. Risk assessment 

Significantly more deliberate self-harm was reported in the psychotic depression group, 33% (n=35) compared with 18% (n=13; p<0.01) at presentation and year 1 follow-up (p<0.01). 40% (n=23) of females with psychotic depression had deliberately self-harmed.

At presentation, significantly more patients with schizophrenia were likely to have harmed others (23%; n=17; p<0.01), harmed property (18%; n=13; p<0.05), and used illicit drugs (32%; n=23; p<0.01). Psychotic depression patients had a comparatively low risk of harming others (6%; n=6), harming property (7%; n=7), and illicit drug use (12%; n=13).

3.7. Changes in diagnosis 

Diagnoses were found to be stable over time in both groups. Of the 105 patients who were initially diagnosed with psychotic depression, 91 (87%) kept the same diagnosis. Of those whose diagnosis changed, 4 changed to a diagnosis of schizophrenia (F20). Of the 73 patients who were diagnosed with schizophrenia at presentation, 67 (92%) kept the diagnosis.

4.1. Summary of the results 

Psychotic depression was more common than schizophrenia in first-episode psychosis, but not in younger people. There were similar hospitalisation rates for both groups. Differences between the groups of days spent in hospital were not significant at the 5% level. However, significantly fewer patients with psychotic depression were admitted under the Mental Health Act. There was no difference in mortality between the two groups.

Concerning treatment, ECT was not often used, despite evidence for its effectiveness. Most patients in both groups received antipsychotic medication, and most patients with psychotic depression also received antidepressant medication.

Patients with psychotic depression had a significantly higher risk of physical health problems and self-harm. Patients with schizophrenia had a significantly higher risk of using illicit drugs, and physically harming others and property.

The ICD-10 diagnoses of psychotic depression and schizophrenia given at first presentation appeared to be relatively stable over time, with only a small percentage changing diagnosis.

4.2. Comparison with the literature 

This study corroborates findings from Baldwin et al. (2005) who found no difference between the incidence of psychotic depression and schizophrenia in rural Ireland. Perhaps surprisingly, the present study revealed a higher incidence of psychotic depression. As in this study, Baldwin et al. (2005) found the mean age at first presentation somewhat higher than for schizophrenia but evident across the entire age range from teens to old age. Although psychotic depression is generally considered to be more common in older adults, previous studies have reported that younger age onset was a more common characteristic of psychotic rather than non-psychotic depression (Lattuada et al., 1999, Thakur et al., 1999). We found psychotic depression to be less common in younger people (under 36 years), with a mean age of 52 years. The mean age for the schizophrenia group was 40.

Very few studies have compared psychotic depression with other psychotic disorders, as opposed to major depressive disorders. Only two previous studies have compared psychotic depression with schizophrenia (Tsuang and Coryell, 1993, Craig et al., 1997). These studies looked at various outcomes, such as psychosocial functioning, employment and psychiatric symptoms and found that patients with psychotic depression had much better outcomes than patients with schizophrenia. We did not measure these outcomes, but if poor outcome includes a high hospitalisation rate then our data corroborated those of Craig et al. (1997). There was no difference in overall hospital days between psychotic depression and schizophrenia. Considering that admission to hospital is the most expensive form of psychiatric treatment (Johnstone and Zolese, 1999), psychotic depression is a very costly disorder, yet it remains severely under-investigated.

The NICE ECT technology appraisal (2003) recommended that ECT should be used in individuals with a severe depressive illness after an adequate trial of other treatment options had proven ineffective and/or when the condition was considered to be potentially life threatening. In the present study, the number of patients with psychotic depression treated with ECT was surprisingly low, considering that many clinicians consider ECT to be the most effective treatment for psychotic depression, particularly for older adults (O'Connor et al., 2001). There is some evidence to suggest that ECT should be used as the primary treatment for psychotic depression (Taylor et al., 2005, Parker et al., 1992, Kroessler, 1985, Petrides et al., 2001, Birkenhager et al., 2005). However, it was difficult to draw conclusions from these studies, due to them comparing ECT treatment with medications in different combinations, dosages and for different periods of time (Rothschild, 2003). There are no RCTs in which ECT was compared directly with pharmacological treatment for psychotic depression (Wijkstra et al., 2005). The present study supported the view that ECT is effective for psychotic depression. According to the Northumberland ECT Treatment Centre, which uses clinical assessment and patient evaluation forms to measure the success of the treatment, all 18 patients who had received ECT were reported to respond well.

It is recognised that patients with mental health problems, particularly schizophrenia, are at increased risk of physical health problems (Mitchell and Malone, 2006). Compared with non-psychotic depression, psychotic depression was characterised by worse physical health (Simpson et al., 1999). No research was found that compared the risk of physical health problems in psychotic depression with other psychotic disorders, but this study indicated that the risk of physical health problems was higher in psychotic depression than in schizophrenia. This association has not previously been reported and the cause is unknown. Substance misuse is known to be a factor in the physical health problems of patients with schizophrenia (Santhouse and Holloway, 1999), as are environmental risk factors including violence (Swanson et al., 2006). Compared with schizophrenia, psychotic depression had a low risk of substance misuse, harm to others and property, but a higher risk of physical health problems. More research into the reasons behind the high incidence of physical health problems in psychotic depression is urgently needed.

There is very little published work on self-harm in psychotic depression or schizophrenia. In affective psychosis and schizophrenia research interest has focused on suicide to identify risk factors, and there has been far less emphasis on acts of self-harm (Symonds et al., 2006, Skegg, 2005). The links between depression and self-harm are clearer but not fully understood (Gladstone et al., 2004, Oquendo et al., 2005). The present study found a high occurrence of self-harm in the psychotic depression group; effective treatment for this disorder might be expected to protect against this.

It is sometimes thought that consultant psychiatrists are unwilling to diagnose a first psychotic episode as schizophrenia due to the negative connotations associated with that diagnosis. It has also been argued that psychotic depression is frequently misdiagnosed (Schatzberg, 2003). When the psychosis is subtle or concealed, this can lead to a misdiagnosis of non-psychotic depression, and when the psychosis is prominent, it can lead to a misdiagnosis of schizophrenia spectrum disorders (Rothschild et al., 2006). Previous studies have reported findings suggesting that psychotic depression in younger adults may be the first episode of a bipolar disorder (Strober and Carlson, 1982, Akiskal et al., 1983). This study found diagnoses of schizophrenia and psychotic depression to be relatively stable over time. 87% of patients with psychotic depression kept their diagnosis, and no patients developed a diagnosis of bipolar disorder. However, it is not possible to draw any conclusions due to the small number of patients with psychotic depression under the age of 36 years (n=31).

4.3. Implications for the future 

Patients with psychotic depression had as many hospital admission days as schizophrenia patients, but perhaps this group has been overlooked because they have had shorter hospital stays. Psychotic depression is a costly condition. An adult acute inpatient day in Northumberland currently costs £258. Patients with psychotic depression had 10,025 hospital days, which means that between October 1998 and October 2006. Psychotic depression cost the NHS in the region of £2.6 million for adult inpatient days. This compares to about £2.8 million for patients with schizophrenia. This high cost further illustrates the need for greater awareness and more energetic treatment of this condition.

Psychotic depression is generally thought to be more common in older people, but this study confirms a previous study (Baldwin et al., 2005) in finding that it is spread across the age ranges. This has implications for Early Intervention in Psychosis teams, who should be looking out for psychotic depression.

ECT was not often administered, perhaps for a variety of reasons. ECT is controversial but clinicians should be aware of the evidence supporting it and of the high risk of self-harm and physical health problems in this group. Regular physical health checks should be part of a treatment protocol. There are no clear guidelines for treating psychotic depression. It is still not known whether combination therapy with antipsychotic and antidepressant medication is the best treatment. The gaps in the knowledge base together with the high treatment costs of psychotic depression reported here, illustrate the need for further research leading to evidence based treatment guidance.

4.4. Strengths and weaknesses of this analysis 

Many studies exclude patients using arbitrary upper age cut-offs. For this study all patients with a first-episode psychotic depression or schizophrenia 16 years and over were included. A unique strength of this study is that it utilised 8 years worth of prospective follow-up data, and data were gathered from a defined catchment area with a relatively stable population.

However, the PACE methodology has limitations. A major limitation is gaps in the clinical records. A number of patients stopped contact with secondary care services within the first year, either on a planned or unplanned basis. There was no way of identifying outpatients with first-episode psychosis, apart from via consultant psychiatrists, which may have led to an under-representation of outpatients.

4.5. Conclusions 

Psychotic depression is a surprisingly common disorder, with a similar hospital admission rate to schizophrenia. The NHS and local authorities spent around £4.5 billion on adult mental health services in 2004/2005 (Mental Health Strategies, 2005). It has been estimated that improved mental health care could actually save the government as much as £3.1 billion per year (National Institute for Mental Health in England, 2005). Although there is now clear guidance on the treatment of schizophrenia, non-psychotic depression and a growing number of other disorders, there is a marked shortage of research and very little guidance on the treatment of psychotic depression. Only by increasing our knowledge of psychotic depression, will we be able to formulate ‘menus of treatment’ to enable us to better treat the disorder and to become more cost-effective.

These findings should be interpreted with caution due to limitations of the methodology, and this study has highlighted the necessity of further research and replication of these findings. There is a clear need for consensus and guidance on best clinical practice in psychotic depression.