Transcript SAPC - QResearch

Risk of malignancy in
patients with mental health
problems
Julia Hippisley-Cox
Yana Vinogradova
Carol Coupland
Chris Parker
SAPC, Keele
July 2006
Aims of presentation
• Overall
– Present analysis of study examining risk of
cancer in patients with mental health
problems
Acknowledgements
Co-author
– Julia Hippisley-Cox
– Carol Coupland
– Chris Parker
QRESEARCH team
– Mike Pringle
– Mike Heaps
– Gavin Langford
– David Stables
– EMIS and EMIS practices
Funding body
Funded by Disability Rights
Commission
Part of programme of work looking at
health inequalities & outcomes for
patients with serious mental health
problems
• Schizophrenia
• Manic depression
Background
• No consistency in literature about
cancer incidence in patients with
mental health problems
– Protective effect of schizophrenia
– Increased cancer incidence and mortality
• Fundamental differences in
methodology
– Cohorts are often restricted to hospital
patients
– No control for confounding factors
Study design & setting
• Nested case control study
• Study period Jan 1995-July 2005
• Separate sets of matched cases and
controls for each cancer
– Breast
– Colorectal
– Gastroesophageal
– Prostate
– Respiratory
Cases & controls
CASES
• 1st ever record of cancer during 10 year
study period
CONTROLS
• 5 controls matched by
– Age
– Sex
– Practice
– Calendar year
Study population:
QRESEARCH database
• Currently largest database in the UK
• 537 UK practices
• > 6 practices in every Strategic Health
Authority (administrative area)
• > 9 million patients including those who
died, left and still registered
• > 30 million person years of observation
Data source:
QRESEARCH database
•
•
•
•
•
Patient level consolidated database
Anonymised data
Longitudinal data for 15+ years
Derived from GP clinical records
Validated against external and internal
measures
• Industry independent
Statistical analysis
• Conditional logistic regression
• Odds ratios + 95% CI
• Unadjusted & adjusted
Sample:
Breast
Colorectal
11,275 cases
47,924
incident cases
of cancer
1995/2005
Prostate
Respiratory
8,356 cases
Gastrooesophageal
3,854 cases
10,190 cases
7,506 cases
49 S
47 MD
40 S
19 MD
14 S
9 MD
13 S
16 MD
23 S
35 MD
53,621 controls
40,010 controls
18,477 controls
48,748 controls
35,981 controls
164 S
173 MD
101 S
90 MD
53 S
41 MD
116 S
93 MD
148 S
103 MD
Proportions of patients with basic characteristics
80
70
64
60
57
59
60
58
52
50
46
42
40
36
34
31
30
17
20
19
19
14
10
0
older 65
females
schizophrenia
© QRESEARCH 2006, version 8
deprived areas
manic depression
obese
smokers
no mental health problems
Confounding factors
Townsend score
• quintiles
Body mass index
•
•
•
•
Less than 25 kg/m2
25 to 29.9 kg/m2
30 kg/m2 or more
BMI not recorded
Smoking status
• Non-smoker
• Smoker
• not recorded
Medications
•
•
•
•
•
NSAIDs
Statins
Cox2 inhibitors
Aspirin
Hormone replaced
therapy
• Contraceptive pill
• Antipsychotic drugs
• Antidepressants
Unadjusted risk of cancer for patients with mental health problems
Colorectal
Gastroesoph
Prostate
Respiratory
100
150
200
250
300
Breast
0
50
100%
S
M
risk
© QRESEARCH 2006
S
M
S
M
S
M
S
M
S - Schizophrenia, M - Manic depression
lower/upper limits of 95% confidence interval
Adjusted risk of cancer for patients with mental health problems
Colorectal
Gastroesoph
Prostate
Respiratory
100
150
200
250
300
Breast
0
50
100%
S
M
risk
© QRESEARCH 2006
S
M
S
M
S
M
S
M
S - Schizophrenia, M - Manic depression
lower/upper limits of 95% confidence interval
Methodological strengths
• Large sample size and representative
population
• Data electronically collected –unlikely
misclassification bias
• Data collected before the diagnosis no recall bias
• Excluded diagnoses of mental health
problems 12 months prior to cancer
diagnosis – no information bias
Risk of cancer compared to general
population
Cancers
Schizophrenia
Manic
depression
Breast
+ 42%*
+ 26%
Colorectal
+ 90%**
+ 1%
Gastroesaphageal
+ 9%
0%
Prostate
- 41%
- 16%
Respiratory
- 46%*
+ 18%
* p < 0.05
** p < 0.01
Conclusions
• Risk of certain cancers differ among
people with schizophrenia compared
with general population
• Patients with manic depression have
similar cancer risk compared with
general population