This document summarizes a study of patients accessing CASCAID, a community mental health service for patients with HIV and mental illness in South-East London. The study analyzed anonymized clinical records from January 2007 to May 2015. It found that of the 1696 adult patients seen, 1035 had a coded psychiatric diagnosis. The cohort was predominantly male (68%) and either black/African (52%) or white (33%). Depressive and adjustment disorders were the most commonly coded diagnoses. The study aims to further analyze free-text clinical notes to gain more diagnostic and demographic information about this population.

1. HIV In A South London Mental Health Population
David O’Flynn1, Rosie Mayston2, Bartolomiej Pliszka1,2, Sophie Candfield3, Chris Taylor3
1. South London and Maudsley NHS Foundation Trust; 2. Institute of Psychiatry 3. Kings College Hospital NHS Trust
AIM: To describe the cohort of people accessing a
community mental health service for patients with HIV
and mental illness in South-East London.
BACKGROUND: We have developed a strong
multidisciplinary collaborative team to explore the
important but neglected area of the interface between
mental health and HIV.
CASCAID is a unique community-based service in
South East London, which provides care for people
with HIV who have mental health issues. They receive
referrals from medical teams, inpatient and community
mental health teams. It does not accept referrals for
patients who are homeless.
Our team has access to a new and extensive database
of anonymised patient records, and to start our
research we have selected the CASCAID cohort from
these records and present their demographics.
AIMS AND BACKGROUND RESULTS
METHOD
Clinical Record Interactive Search (CRIS) provides
authorised researchers with regulated access to
anonymised information extracted from the South
London and Maudsley NHS Foundation Trust
(SLaM) electronic clinical records. Data presented
here describes adult CASCAID patients accessing
CASCAID from January 2007 to May 2015. These
patients are a subset of those accessing services for
HIV and mental illness in South-East London, which
has the highest rates of HIV infection in the UK
(Public Health England 2014).
Where data was not available for a specific criterion,
the patient was excluded from that analysis
CONCLUSIONS
Baseline analysis of coded diagnoses in the
CASCAID HIV/mentally ill population showed a
largely male (67%), and predominantly black
or African population. Most males were
white, and most females were black.
There was a large variety of coded ICD-10
diagnoses, however, this data was in general
poorly recorded in the coded information.
Depressive and adjustment disorders were
the most commonly coded diagnoses
Figure 5: Primary coded (ICD-10) psychiatric
diagnosis where available (N=611).
Figure1 : HIV in South East London
0
100
200
300
400
500
600
700
800
18 - 24 25 - 34 35 - 45 46 - 54 55 - 64 65 - 76
Figure 3: Age profile of adult patients under
CASCAID (N=1696)
We found 1696 adult patients who had been seen
by CASCAID over this time. 1035 had a diagnosis
coded by the CASCAID team. The primary
diagnosis was a psychiatric diagnosis in 611
patients.
Figure 2: Basics of CRIS database usage.
32%
68%
Female
Male
Figure 4: Sex of adult patients under CASCAID
(n=1696)
Figure 5: Ethnicity of adults under CASCAID
(n=1488)
Figure 6: Ethnicity of adults under CASCAID,
by sex, as a percentage of total adults under
CASCAID (n=1488)
8%
7%
6%
5%
5%
4%
4%
3%
2%
2%
2%
52%
F43.2 - Adjustment disorders
F43 - Reaction to severe stress, and adjustment disorders
F32.1 - Moderate depressive episode
F41.2 - Mixed anxiety and depressive disorder
F32.0 - Mild depressive episode
Z71.1 - Person with feared complaint in whom no diagnosis is made
F99 - Mental disorder, not otherwise specified
F00-F99 - Mental and behavioural disorders
F33.1 - Recurrent depressive disorder, current episode moderate
F02.4 - Dementia in human immunodef virus [HIV] disease
F32 - Depressive episode
Other diagnoses
DISCUSSION
Coding was only available for limited
demographic information, and this was
incomplete in some cases. Information on
sexual orientation was, for example, not
collected.
For this reason our group intends to
interrogate the “free text” data on CRIS of
patient encounters using Natural Language
Programming techniques to gain diagnostic,
demographic, and comorbidity data. We
believe that working with CRIS data will
enable us to construct a large cohort of
people living with HIV and SMI and will
therefore provide an important contribution
to the evidence base of this uncommon but
potentially burdensome comorbidity.
38%
33%
18%
11%
African or other
black background
British
Other White
background
Other
24
4 2 3
33
14
29
16
8
67
0
10
20
30
40
50
60
70
80
90
100
African or
other black
background
British Other
White
background
Other Total
Male
Female
, Shaz Alikhan
1
D.Chandra
1