Contents

Research within CPFT

Research is critical to the long-term success of CPFT. Together with our research partners (including the University of Cambridge, the University of East Anglia, and Cambridge University Health Partners), we undertake research that contributes to the development of medical knowledge and supports innovation in mental health services. Please explore this section of the web site to learn more about research in CPFT.

The CPFT Research Database

 

Introduction to the CPFT Research Database

Dr Rudolf Cardinal. Click above or here to hear Rudolf explain how the CPFT Research Database works (2014).

What is the CPFT Research Database?

It is a database of anonymised medical records. CPFT’s electronic clinical records are put through a software tool called CRATE (Clinical Records Anonymisation and Text Extraction), which removes information that might identify a patient.

You can opt out of your anonymous data being used for research (see Opting Out below). If you opt out, your data will not be held in the database or used for data linkage (see Data Linkage below).

Anonymising clinical records to create the secure CPFT Research Database
Figure 1. Anonymising clinical records to create the secure CPFT Research Database.

What does the database help to achieve?

The database is invaluable as researchers carry out their work into what causes conditions such as dementia, and which treatments are best for dementia and other mental health conditions.

How is the Research Database used?

The database allows research using this anonymous information. For example, researchers might use the database to see which treatments help the most in schizophrenia, or how often patients get side effects such as weight gain from a particular medicine, or how well different kinds of psychological therapy work for different conditions. In this kind of research, the researchers do not know the identity of any patient.

The database speeds up research, and makes research possible that would otherwise be impossible. It also allows research to be conducted anonymously when in years gone by the same research would have needed research clinicians to see paper case notes.

Obviously, for some kinds of research, the researchers and patients must meet face to face. The database can also help this process. Researchers can search the anonymous records for patients that might be suitable for their study, without knowing their identity. If and only if the service user gives their specific permission for their details to be released to the researchers, a special process (outside the database) allows this to happen. (See “How can I choose to be contacted about research?” below.)

Here, the database helps people participate in research. It means that when people get asked about research studies, those studies are very likely to be relevant to them. It helps researchers to recruit participants faster and to find the answers to important medical questions faster.

Is the database ethical? Who approved it?

  • The database was developed with extensive involvement from patient/service user representatives, and is overseen by a Research Database Oversight Committee that includes patients and doctors.
  • The database has been approved by an independent NHS Research Ethics Committee specialising in databases.
  • It has also been approved by CPFT’s Research and Development department and its Board of Directors.
  • The database originally used the CRIS system developed at the NIHR Maudsley Biomedical Research Centre by the South London and Maudsley NHS Trust (SLaM) and the Institute of Psychiatry, Psychology and Neuroscience (IoPPN) at King’s College London (KCL). SLaM’s equivalent database has been running since about 2008. The UK National Information Governance Board (replaced in 2013 by the Care Quality Commission) hailed this system as a model for improving research while ensuring patient anonymity. CPFT now uses the CRATE system developed in CPFT and the University of Cambridge Department of Psychiatry.

Detail of external approvals:

How secure is the information?

The overriding principle is: nobody except the care-giving team can see confidential patient information (identity plus medical information) without the patient’s specific consent. Researchers never see any patient-identifiable information without the patient’s specific consent.

  • Patient-identifiable information is removed before information is stored in the Research Database.
  • The database is held within CPFT’s secure computing environment, within the NHS. No information ever leaves this environment except summary information about large groups of people, from which no individual could be identified.
  • In some circumstances, CPFT’s Clinical Data Linkage Service (see CDLS below) uses identity information (e.g. name, NHS number) temporarily for linking two data sources. This requires special legal permission. This process is performed by CPFT administrative staff, or sometimes by specific secure UK national services. They never see identifiable medical information. In practice, computers process the data and administrative staff supervise them. The identity information is removed before researchers see the data. Researchers never see identifiable information without the patient’s consent.
  • For a few special linkages via CPFT’s Clinical Data Linkage Service (see CDLS below), de-identified CPFT data is analysed in UK national secure computing facilities (such as the Office for National Statistics [ONS] Secure Research Service [SRS]). This is done under specific approvals by the NHS and to security standards that meet or exceed those of the NHS.
  • All researchers using the database must be working within CPFT – either as clinicians, or as affiliated researchers approved by CPFT to work within CPFT. For example, our researchers include psychiatrists and psychologists working for CPFT, the University of Cambridge, or both.
  • All research projects using the database must be approved by CPFT’s Research Database Oversight Committee. Projects need to meet the highest of standards for access to the database to be permitted.
  • Linkages that involve the temporary handling of identity information must have special approval under Section 251 of the NHS Act 2006, and must apply the NHS National Data Opt-Out (as well as CPFT’s local opt-out systems).
    • The NHS National Data Opt-Out applies to the use of confidential patient information, for purposes beyond an individual’s care and treatment, where the legal basis to use the data is section 251 of the NHS Act 2006. (See “National Data Opt-out: Factsheet 2 – When it applies”, 25 May 2018, or alternatively NHS Digital’s information page.) “Confidential patient information” is identifiable information relating to physical or mental health (see section 251, subsections 10–11, of the NHS Act 2006).
  • All projects involving contact with patients must also have their own independent approval from an NHS Research Ethics Committee.
  • Access to anonymous data for researchers is strictly controlled.
  • Access to patients by researchers is even more strictly controlled (see How can I choose to be contacted about research? below).

Patients decide on a colour flag, which determines whether researchers may contact them, summarised below. If they’ve not yet expressed a preference, the “yellow” route is taken.

CPFT Research Database consent-for-contact traffic light system
Figure 2. CPFT’s traffic-light “consent for contact” system.

If you want to choose or change your traffic-light preference, talk to your clinician or care coordinator, or contact us (see “Contacting us” below).

Is it an opt-in or an opt-out system?

1. The anonymous database is opt-out. If you have had information about you recorded electronically at CPFT since 2005, your information will have contributed to this database, but with identifying information (names, addresses, exact dates of birth, NHS and CPFT numbers, names of family members, and so on) removed.

Thank you! If you are one of these people, you are anonymously helping to save lives.

If you want to opt out, talk to your clinician or care coordinator, or contact us (see “Contacting us” below).

2. Having researchers contact you, and having them see your identifiable medical records, is always opt-in. Researchers outside your clinical care-giving team will never be allowed to contact you (or even know who you are) via the database unless you have given your specific permission.

How many patients’ details are in the database?

As of October 2020, the database contains anonymised information derived from about 260,000 people who had care from CPFT.

Will pharmaceutical or insurance companies be able to access the database?

No.

Sometimes pharmaceutical companies conduct clinical trials in CPFT and other parts of the NHS. These trials are very strictly regulated by the NHS. So the pharmaceutical companies know which parts of the NHS are best to work with for a particular trial, they sometimes ask CPFT: “How many patients in your service have disease X?” CPFT sometimes uses information from the Research Database to answer these questions. CPFT’s answer is a single number, such as “2,400” or “Fewer than 10”. The companies never get to see or use the Research Database. If the Database is subsequently used to find patients anonymously who might be suitable for research, the Database is again used by CPFT researchers, not by external companies.

Are the details in the CPFT database given or sold to third parties?

No.

Via the CPFT Clinical Data Linkage Service, identity information (e.g. names or NHS numbers) – but not medical information – is sometimes passed to other UK national bodies (e.g. NHS Digital) so we can link other records to CPFT data. This identity information is then deleted (see Data Linkage below). This process needs special legal permissions. Similarly, via CPFT’s Clinical Data Linkage Service, anonymous CPFT information is sometimes analysed in UK national high-security data centres, but only by CPFT-approved researchers.

CPFT does not give or sell information from the Research Database to other parties.

How much has this cost?

About £350,000 to set up – of which £300,000 came from the UK National Institute of Health Research (NIHR, part of the NHS and funded by the Department of Health) and about £50,000 from CPFT.

Some funding was as part of the D-CRIS project, which was a 2014 research collaboration between five NHS mental health Trusts with their partner universities. It was funded by the NIHR to create a world-leading resource for collaborative dementia research. The five were: Cambridgeshire and Peterborough NHS Foundation Trust (with the University of Cambridge and the NIHR Cambridge Biomedical Research Centre); Camden and Islington NHS Foundation Trust (with University College London); Oxford Health NHS Foundation Trust (with the University of Oxford); South London and Maudsley NHS Foundation Trust (with the Institute of Psychiatry, King’s College London), hosting the NIHR Mental Health Biomedical Research Centre and Dementia Unit; West London NHS Trust (with Imperial College London).

What future developments might there be?

We may ask for permission to link CPFT data to other data sources via the CPFT Clinical Data Linkage Service. This might include acute hospital data, or data from primary care (GPs).

The University of Cambridge is developing supercomputing facilities certified to very high security standards (ISO-27001). Advanced computing may help in developing new ways to diagnose and treat diseases (for example, by analysing brain scans). CPFT does not have this kind of computer facility. We may ask for permission to analyse CPFT data (without information that would identify people) in a special “NHS zone” of the University of Cambridge, to help develop new methods for use within the NHS. Access to the information would be restricted in ways approved by the NHS.

We might join other international health care organizations in supporting a “federated” query system, so that researchers can analyse data from all over the UK or all around the world, without ever having access to information about individual patients. (A “federated” or “distributed” query is one where computers send out requests to lots of sites. Each site’s computer assembles part of the answer, which does not contain individual patient information. The part-answers are then sent back to the original site, which puts them together to make a whole answer.) This could lead to advances in research. We would never make individual patient data available in this way, just summaries of information from large numbers of people.

Data linkage

 

Introduction to data linkage

Linking data means joining two or more different databases, so that information about a person in database 1 can be connected to information about the same person in database 2. This requires some shared information. For example, someone’s GP record and their hospital record will share an NHS number. Linkage helps to improve the quality of information, and means that researchers can look at patients’ healthcare in more detail. For example, linking mental health data to physical health data lets researchers look at how physical health and mental health are related (such as major physical illnesses that are commoner in schizophrenia, or the relationships between depression and heart disease).

Click above or here to watch a video (“Data linkages: explore the evolution of healthcare records in research”, 2020, by our colleagues at King's College London) explaining data linkage for healthcare records. Some background information is also provided here.

The CPFT Clinical Data Linkage Service (CDLS)

We have set up a Clinical Data Linkage Service (CDLS). This lets us join or link information (“data”) from the CPFT Research Database with other data. This is done within a secure “safe haven”. In the NHS, Safe Havens are set up to link confidential patient information in a way that guarantees patients’ legal and ethical rights. The CDLS has been approved by an independent NHS Research Ethics Committee.

Although data is sometimes linked using “identifiers” such as the NHS number, these identifiers are then all removed. The information is fully anonymised (de-identified) before any researchers see it. Sometimes, data can be linked without using any directly identifying information.

The CDLS has linked the CPFT Research Database with other databases. We describe these below.

This picture gives an overview of the CDLS:

Overview of the CPFT Clinical Data Linkage Service
Figure 3. Overview of the CPFT Clinical Data Linkage Service.

All linked data is de-identified. Even though the researchers can’t identify you from your data, you can opt out (stop your data being used in this way) if you want.

As well as all data being de-identified, researchers will only be able to see data that they need to for their project, even though the full CPFT records will likely hold other information too. For instance, if a project is looking at death certificate information, the researchers will not be able to see any information that is not relevant to that project (such as notes made during reviews of a patient’s mental state).

The “traffic light” system (being contacted about research) is not used for linked data.

The data is linked in secure computer systems, not on paper. It is linked by computer. The computers are supervised by technical experts in CPFT and the organizations providing linked data (not by researchers, or by clinical staff involved in an individual’s care).

Below, we list and explain every data source that is linked to CPFT data via the CDLS. At present these are:

CPFTRD – NHS Hospital Episode Statistics (HES)

We link CPFT information to central NHS information on hospital attendances and admissions. Researchers do not see the identities of any patient. You can opt out in several ways. The work has been approved by an independent NHS ethics committee. This section explains why we do this, and how the process works. The picture below gives a summary.

Previous research has found that physical health is sometimes worse in people with mental health problems. Also, some physical illnesses increase the chance of mental health problems. To understand this relationship, and to improve physical health and life expectancy for people with mental health problems, researchers need information about both mental and physical health conditions.

Hospital Episode Statistics (HES) is a database held by NHS Digital. (NHS Digital is the information technology branch of NHS England.) It includes all accident and emergency, hospital admissions, and outpatient visits which occur in all hospitals throughout England. The CDLS has linked the CPFT Research Database (CPFTRD) to HES. This linkage includes important clinical information such as diagnoses, operations, or the speciality of the treating clinician; demographic information such as age, gender, and ethnicity; and administrative data such as methods of admission and discharge.

The CDLS holds de-identified (anonymised) HES data on all CPFT service users who have attended hospitals anywhere in England. The CDLS also holds anonymised HES data on the residents of Cambridgeshire and Peterborough (the CPFT catchment area). This lets researchers compare physical illnesses and hospital use of CPFT service users to the general population. This linkage also lets researchers investigate risks for physical illness in particular patient groups, or look at how pre-existing physical health problems may affect the outcomes of mental health treatments. The linkage between CPFTRD and HES is important. We hope the information on physical health from HES will let us find ways to improve care for people with mental health problems.

To link CPFTRD data with HES, we send identifiable information for CPFT patients (NHS number, name, date of birth, and postcode) to NHS Digital. NHS Digital can then find information on hospital records from England for the relevant people. This linkage is legal with the Health Research Authority’s support under Section 251 of the NHS Act 2006. This permission allows temporary use of identifiers for accurate linkage in a secure computing environment. Once NHS Digital has linked the records, all personal information is destroyed (NHS number, name, date of birth, and postcode) and the linked records are sent securely back to the CDLS. The data is fully de-identified before any researchers are given access to it. The process is drawn below.

Linking the CPFT Research Database to NHS Hospital Episode Statistics (HES)
Figure 4. Linking the CPFT Research Database to NHS Hospital Episode Statistics (HES).

Further details about HES can be found on the NHS Digital website.

The CDLS does not use data from people who have told CPFT they wish to opt out, or who have opted out nationally via the NHS National Data Opt-Out.

If you have any questions, or would like to tell us that you do not wish to have your records in the CPFTRD or linked with other databases, please contact us (see “Contacting us” below).

CPFTRD – UK Office for National Statistics (ONS) Mortality

We link CPFT information to UK information on causes of death. Researchers do not see the identities of any patient. The work has been approved by an independent NHS ethics committee. This section explains why we do this, and how the process works. The picture below gives a summary.

Previous research has found that people with serious mental disorders generally have a lower life expectancy than other people. Linking the CPFT Research Database (CPFTRD) to information causes of death (which is not held by CPFT itself) is important. It helps researchers understand why this might be, and provides information that we hope will help in improving the life expectancy of people with mental health problems.

The UK Office for National Statistics (ONS) collects information on causes of death from people’s death certificates. This information is held by NHS Digital. The CDLS has linked the CPFTRD with the ONS Mortality data. This includes date-of-death and cause-of-death information for all CPFT patients who had died before the most recent data linkage took place. This linkage allows researchers to investigate the relationship between mental health and cause of death. The CDLS also holds anonymised mortality data on the residents of Cambridgeshire and Peterborough (the CPFT catchment area). This lets researchers compare causes of death of CPFT service users to the general population.

To link CPFTRD data with ONS Mortality data, we send identifiable information for CPFT patients (NHS number, name, date of birth, and postcode) to NHS Digital. NHS Digital can then find mortality records from England and Wales (provided by the ONS) for the relevant people. This linkage is legal with the Health Research Authority’s support under Section 251 of the NHS Act 2006. This permission allows temporary use of identifiers for accurate linkage in a secure computing environment. Once NHS Digital has linked the records, all personal information is destroyed (NHS number, name, date of birth, and postcode) and the linked records are sent securely back to the CDLS. The data is fully de-identified before any researchers are given access to it. The process is drawn below.

Linking the CPFT Research Database to ONS mortality data
Figure 5. Linking the CPFT Research Database to ONS mortality data.

Further details about the ONS Mortality Data can be found on the NHS Digital website.

The CDLS does not use data from people who have told CPFT they wish to opt out, or who have opted out nationally via the NHS National Data Opt-Out.

If you have any questions, or would like to tell us that you do not wish to have your records in the CPFTRD or linked with other databases, please contact us (see “Contacting us” below).

CPFTRD – NHS National Cancer Registration and Analysis Service (NCRAS)

We link CPFT information to central NHS information on cancer. Researchers do not see the identities of any patient. You can opt out in several ways. The work has been approved by an independent NHS ethics committee. This section explains why we do this, and how the process works.The picture below gives a summary.

There is evidence that cancer survival rates are lower in people with severe mental disorders, suggesting that research should focus on their care.

The National Cancer Registration and Analysis Service (NCRAS) is run by Public Health England (PHE). It collects information about every patient diagnosed with a malignant tumour in England, in order to support public health, service monitoring, and research. The NCRAS collects information on over 300,000 cases of cancer every year, including detailed data about the type of cancer, how advanced it is, and the treatment the patient receives. In collaboration with PHE, the CDLS has linked the CPFT Research Database (CPFTRD) to cancer data from the NCRAS along with information on hospital visits from HES, and mortality data.

The CDLS holds de-identified cancer data on all CPFT service users who have received a diagnosis of cancer anywhere in England. In addition, the CDLS also holds anonymised cancer data on the residents of Cambridgeshire and Peterborough (CPFT’s catchment area), as a “control” or comparison group. This will enable us to investigate in more detail the care received by people with mental disorders after a cancer diagnosis and the outcomes people experience, in order to investigate whether differences in these might explain differences in cancer survival. By generating information on cancer detection and care for people with mental health problems in the UK, this project aims to enhance mental health care among cancer patients – whether or not they have an existing mental health problem – and improve the quality of end-of-life care for people with both cancer and mental health problems.

To link CPFTRD data with NCRAS data, we send research “tags” (encrypted NHS numbers) to PHE. PHE can find cancer registry data for people who have the same research “tag”. They send de-identified information securely back to the CDLS, where it is linked to CPFT data using the tag. Researchers do not learn anyone’s identity. The process is drawn below.

Linking the CPFT Research Database to NHS national cancer data
Figure 6. Linking the CPFT Research Database to NHS national cancer data.

Further details about the national cancer registry can be found on the NCRAS website.

The CDLS does not use data from people who have told CPFT they wish to opt out, or who have opted out nationally via the NHS National Data Opt-Out.

If you have any questions, or would like to tell us that you do not wish to have your records in the CPFTRD or linked with other databases, please contact us (see “Contacting us” below).

CPFTRD – UK National Pupil Database (NPD)

We link CPFT information to information from a UK education database. Researchers do not see the identities of any patient. You can opt out in several ways. The work has been approved by an independent NHS ethics committee. This section explains why we do this, and how the process works. The picture below gives a summary.

One in eight 5–19 year olds had a mental disorder in 2017 and this rate is increasing. Educational factors affect mental health, and mental ill health affects education. It’s therefore important to do research on the links between education and mental health. Please note that this linkage is about children’s mental health, not parents’ mental health. All records are de-identified (no identifiable information about individuals is seen by researchers, or shared with NHS clinicians, or with schools). The data is managed in an extra-high-secure national UK research environment.

The UK Department for Education (DfE) runs the National Pupil Database (NPD). This contains information from schools and colleges in England, including exam results; special educational needs; absences; free school meals eligibility; demographic information such as age, gender, and ethnicity; and administrative information such as school type. The DfE make it available for approved research purposes.

Linking this data to the CPFT Research Database will allow researchers to investigate links between mental health and educational outcome. This might go in either direction: education problems might affect mental health, and mental health problems might affect education. The CDLS holds de-identified data from the Department for Education National Pupil Database linked with de-identified Child and Adolescent Mental Health Service (CAMHS) CPFT records. It also holds data from anonymised control sample of school-age children who were not patients of CPFT but who were at school and lived in Cambridgeshire and Peterborough (CPFT’s catchment area) between 2012 and the most recent linkage.

In order to link CPFTRD data with the NPD, we send identifiable information for a subset of CPFT patients (name, date of birth, and postcode) to the DfE. No medical information is sent to the DfE. The DfE can then find information from the National Pupil Database for the relevant people. This linkage is legal with the Health Research Authority’s support under Section 251 of the NHS Act 2006. This permission allows temporary use of identifiers for accurate linkage in a secure computing environment. Once the DfE has linked the records, all personal information is destroyed (NHS number, name, date of birth, and postcode). The linked records are sent to the UK Secure Research Service, hosted by the UK Office for National Statistics (ONS). CPFT sends relevant research data to the Secure Research Service, where it joins the data from the NPD. This data is fully de-identified before any researchers are given access to it. The process is drawn below.

Linking the CPFT Research Database to National Pupil Database data
Figure 7. Linking the CPFT Research Database to National Pupil Database data.

This linkage will provide evidence to help families, treating clinicians, and health policy makers understand the risk factors for childhood mental illnesses such as autism, attention-deficit/hyperactivity disorders, psychosis, obsessive–compulsive disorders, and depression. It will also provide evidence to help clinicians and families understand the impact of childhood mental health on educational outcomes. Families who seek child and adolescent mental health care want their child to get back on track at school and improve their education performance. This linkage will enable research into how educational risk factors affect childhood mental health and outcomes.

Further details about the National Pupil Database are on the UK Data Service web site.

The CDLS does not use data from people who have told CPFT they wish to opt out, or who have opted out nationally via the NHS National Data Opt-Out.

If you have any questions, or would like to tell us that you do not wish to have your records in the CPFTRD or linked with other databases, please contact us (see “Contacting us” below).

How can I find out more?

For more information on the CPFT Research Database, please see:

There is also:

For more information on the ways in which CPFT looks after your information, please see:

Contacting us about the CPFT Research Database

If you

  • have any questions,
  • want to tell us your traffic-light preference,
  • or would like to tell us that you do not wish to have your records in the CPFT Research Database or linked with other databases (to opt out),

then you can:

  • e-mail research.database@cpft.nhs.uk;
  • write to The CPFT Research Database Manager, FREEPOST CPFT RESEARCH DATABASE MANAGER;
  • phone 01223 219400 and ask for the Research Database Manager;
  • or speak to your care coordinator or other clinician.

The CPFT Research Database manager is Jonathan Lewis.

The page was last updated on 22 October 2020 by aji.kuriakose.

feedback

Cambridgeshire and Peterborough NHS Foundation Trust
Elizabeth House, Fulbourn Hospital
Cambridge, CB21 5EF

T 01223 219400 (open 8:30am to 5pm)
F 01480 398501

IIP-logo-reversed Mindful-employer NHS-Employers-web Triangle-of-Care-member Twitter Logo Youtube Logo Facebook Logo Instagram Logo Wordpress Logo

IP: 18.234.255.5