John McGrath Sukanta Saha Joy Welham David Chant

1
John McGrathSukanta SahaJoy WelhamDavid Chant
A systematic review of the prevalence of
schizophrenia
2
Schizophrenia - the most common form of psychosis
is characterised by disorders of cognition (eg
paranoia), affect (mood), communication (thought
disorder) and perception (eg hallucinations)
leading to a loss of contact with reality along
with various forms of impaired behavior.
old Because schizophrenia (a) has an early
onset and (b) has a relapsing or chronic course,
it is a significant public health problem. It
ranks in the top 10 leading causes of disability,
with a burden of disease comparable to cancer and
greater than heart disease.old Understanding
the prevalence of schizophrenia has important
implications for both health service planning and
risk factor epidemiology.new
3
Aims

• To systematically identify and collate studies of
  the prevalence of schizophrenia
• To summarize variation in time, place and person
  by examining the distribution of these estimates
  of prevalence
• To explore factors which may influence prevalence
  estimates

4
Outline

• Research questions
• Methods Systematic reviews
• Key results
• Caveats and Conclusions

5
The prevalence of schizophreniaHypotheses

• Sex difference
• Males gt females
• Migrant status
• Migrants gt native born
• Urbanicity
• Urban born gt rural born

6
Types of prevalence studies

• Core studies
• Sentinel surveys
• Register based studies
• Migrant studies
• Cohort studies
• Other special groups

7
Methods systematic review

• Electronic data search
• Medline, PsychoInfo, Embase, LILAC
• 1965-2001 inclusive
• (schizo OR psycho) AND
• incidence OR prevalence)
• also
• Review article bibliography
• Wrote to authors
• Screen abstract and reviewed papers to cull
  irrelevant citations

8
Rate items and discrete data

• Non-overlapping
• Sex Male, Female
• Overlapping
• Age eg all ages or age 15-54
• Diagnosis eg Catego S or Catego SPO
  clinical
• Diagnostic categories
• eg DSMIV Schizophrenia or
  Schizophrenia Delusional disorder
• Site overlap eg Denmark or Copenhagen
• Epoch overlap eg 1990-92 or 1989-91

9
Data analysis example cumulative distribution
Rate per 100,000
10
Results

• Electronic search 834 potential papers
• Manual checking 249 potential papers
• Letters from 52 authors, who provided an
  additional 41 references
• Of potentially relevant papers, 74 were
  identified from electronic sources

11
Results (2)

• Rates based on 176,056 potentially overlapping
  incident cases
• After review
• 158 were included from 32 countries
• Types of studies
• Core studies 100
• Migrant studies 24
• Cohort studies 23
• Other special groups 14

12
Core Prevalence Studies

13
Sex differences
14
Male female rate ratio

15
Migrant status
16
Urban-rural differences
17
Cohort studies

• Australia (n2)
• Denmark (n3)
• Finland (n5)
• Israel (n2)
• Italy (n1)
• Sweden (n2)
• The Netherlands (n2)
• USA (n3)
• United Kingdom (n3)

18
Other special groups

• over age 65
• twins
• various ethnic and/or religious subgroups
• students
• deaf individuals
• workers in a radiation contamination zone

19
Key findings

• Most of the distributions are data rich
• Variation
• Asymmetrical
• Long upper tail (gt25 rates)
• Median 15.2 (10-90 7.7- 43.0) per 100,000
• Five-fold range within the 10-90 quantiles

20
Key findings (2)

• Males gt females
• Migrants gt native born
• Urban gt mixed urban/rural
• Schizophrenia has a varied and detailed
  epidemiological landscape

21
Caveats

• Comparisons in systematic reviews should be
  planned, based on directional hypotheses
  limited to a reasonable number
• Systematic reviews are best suited to
  hypothesis-generation
• Geographical boundaries are administrative
• Mostly treated prevalence

22
Conclusions
There is a wealth of data available on the
prevalance of schizophrenia. Studies come from
many countries, with many different
methodological features, and conducted over
several decades. The width and skew of the
distributions, and the significant impact of sex,
urbanicity and migrant status on these
distributions, indicate substantial variations in
the prevalence of schizophrenia. Thus these data
may provide leads for further research into risk
factors