Frequently asked questions - researchers

The iCARE Team are able to provide assistance with any issues faced by researchers relating to the functionality of the iCARE Environment or the data provided. Crucially however the iCARE Team will not assist researchers with their analysis or development of any code. Researchers applying for access to data on the iCARE Environment are expected to already possess the skills necessary to handle and analyse large patient datasets.

Approved researchers can access data within the iCARE environment.  To obtain approvals for your research, you will need to complete and submit a Data Access Request Form and present your project to the NIHR Imperial BRC Data Access and Prioritisation Committee. Anyone accessing data will need to meet the criteria for Accredited Researcher Status We will provide support for applicants to complete mandatory Information Governance training and then sign the relevant Terms of Use for Data. All applicants will be expected to have an undergraduate degree (or higher), including a significant proportion of mathematics or statistics, or be able to demonstrate at least three years quantitative research experience and be proficient in using data analysis and programming tools (such as SQL, R-Studio or Python).  

Your Data Access Request should include a well thought-out proposal with careful consideration and specification of your data requirements, such that you should not need to make frequent amendments. However, should an amendment be required, the Data Access and Prioritisation Committee will consider your request. You should make tracked changes to your original Data Access Request form and then email the amended form together with an updated Data Dictionary to Imperial.dataAccessRequest@nhs.net. If there are significant changes to the original request, your may be asked to present the amendment at a Data Access and Prioritisation Committee meeting to provide full justifcation for the proposed changes.

Understanding data structures, including dimensional and fact tables, along with familiarity with various data types such as integers, strings, and dates, is key to handling the heterogeneous dataset you will be working with. Additionally, knowledge of existing clinical coding systems such as ICD-10, SNOMED, and OPCS-4 will be helpful to identify cohort of patients with certain conditions. Experience in a programming language is also desirable, as it will enable you to effectively utilize the available analytical tools in the environment, such as Snowflake, Python, and R.

Slides are not required when presenting your project to the NIHR Imperial BRC Data Access and Prioritisation Committee during a Committee meeting. The meeting’s secretariat will present your written application to the Committee members, and you will be asked to provide an overview of the projects and answer questions raised by members of the Committee.

If there is a legitimate reason why the project analysis cannot take place on the iCARE Environment, it is possible to transfer data to an external secure data environment. This would however require additional information governance sign off and work. The research team will be responsible for working with the ICHT Data Protection Office on providing justification as to why an external environment is required, and completing all the documentation required to extract data from the Trust.

The NIHR Imperial BRC Data Access and Prioritisation Committee has obtained research database ethics approval for the data held within the iCARE High Performance Environment from the Health Research Authority (HRA). This ethics approval covers all research conducted in the iCARE Environment using the ICHT data held in the platform. Any projects approved by the Committee therefore are covered by this ethics approval, meaning that projects do not need to apply for individual ethics approval from the HRA.

If a project intends to link data held within the iCARE High Performance Environment with secondary data, then additional ethics approval will need to be obtained to cover the use of the secondary data for research. This will not be covered by the current research database ethics in place.

Once a project team is ready to submit a paper for publication they can contact the iCARE team who will provided the details of the HRA ethics approval.

You should prepare a three minute lay-friendly summary of your project outlining why it is important to do the work, your aims and objectives, and why you need access to health data. It is essential that your explain your project in simple terms, avoiding medical abbreviations and technical language; we have lay partners who sit on the Data Access and Prioritisation Committee. You do not need to prepare any slides.

Please submit any questions regarding your project proposal via email.

If your project is utilising data held within the iCARE High Performance Environment only, no further project sign-off or ethics approval is needed. If however you intend on linking iCARE data to secondary data sources as part of your project, additional approval may be required. Please contact the iCARE team prior to submitting your project application to discuss this.

If you’re unable to continue with your project, please contact the iCARE Team on Imperial.dataAccessRequest@nhs.net immediately. The project will then be marked as ended, and access to data will be removed.

Yes as long as you are part of a NHS organisation within Northwest London, or an academic institution associated with it, you can apply for access to data. Please not however that your project’s Clinical Sponsor will need to be a clinician who actively treats patients at Imperial College Healthcare NHS Trust.

Once your project is approved, access to data is initially granted for a six-month period. After six months, you will be asked to provide a written update on the project. Questions will focus on what progress has been made up until that point, and if any of the project’s objections or methods have changed since the initial application.

Access will then be granted for an additional six months. If after that the project is still ongoing and continued access to data is required, the project team will be asked to present the project once again during a meeting of the NIHR Imperial BRC Data Access and Prioritisation Committee. During the presentation the project team will be asked to provide an overview of what works has been completed in the last year, and why the project was not completed within the twelve-month period.

If the update to the project proposal is minor, such as the addition of a new researcher to the project team, this can usually be approved as Chair’s Actions. If however there is a significant update to the project, such a change to the main research question or the data required, a re-application will be required. The project team will need to attend a Data Access and Prioritisation Committee meeting to present the updated proposal.

If you are struggling to log into the ICHT Virtual Desktop Infrastructure (VDI) despite having successfully logged in before, there might be an issue with the Trust’s VDI system. The Imperial College Healthcare NHS Trust ICT department manage the VDI, and are best suited to address any issues. The number for the ICT helpdesk is +44(0)20 3311 5555.

Yes. Any aspects of your project that involve manipulating iCARE data will need to be carried out within the secure data environment.  When you have completed your analysis, you can request that tables, figures, etc be removed.

The NIHR Imperial BRC Data Access and Prioritisation Committee is chaired by Sanjay Gautama and Erik Mayer; the ICHT Caldicott Guardian and Transformation Chief Clinical Information Office respectively. It is also made up of senior clinicians from ICHT, research leads from Imperial College London, representatives from the ICHT Data Protection Office and members of the public (lay representatives).

Imperial College London Patient Experience Research Centre (PERC) has several resources on how to involve patients and the public at all stages of the research cycle. There is no ‘one size fits all’ approach to involving people. It will depend on the needs of the people you are involving, the type of project you are undertaking, the time and resources you have available and what you want to achieve. If you have need any support or have any questions, please get in touch with the PERC team: publicinvolvement@imperial.ac.uk

The iCARE High Performance Environment is a closed system. Researches are expected to utilise the analytics tools available within the platform, meaning no external software can be installed.

Please contact imperial.icareresearchersupport@nhs.net. If an R package is available on CRAN and a Python package on Pypi, it can be approved and installed onto Azure Machine Learning on the iCARE Environment.

Researchers can utilise R and Python for analysis, using Azure Machine Learning notebooks. Additionally, Microsoft Excel is available on the iCARE Environment.

• Introduction to Digital Health
  Introduces the field of digital health and the key concepts and definitions in this emerging field. The key topics include Learning Health Systems and Electronic Health Records and various types of digital health technologies to include mobile applications, wearable technologies, health information systems, telehealth, telemedicine, machine learning, artificial intelligence and big data.
• Foundations of Public Health Practice: The Public Health Approach
  Scope, origins and ethics of public health
• Introduction to Statistics & Data Analysis in Public Health
  Introduces types of variables, common distributions, hypothesis testing
• Introduction to Python
  Python Essential Training
• Introduction to Machine Learning NLP
  Applied Text Mining in Python
• Introduction to SQL (YouTube video)
  Full Database Course for Beginners
• Introduction to Pandas
  Introduction to Data Science in Python
• Introduction to Snowflake
  
  
• Snowflake - Documentation
  
  
• Snowflake SQL
• Snowsight quick tour
• Information Governance - Data Security Awareness
  ICHT iCARE Data Researchers
• Introduction to SQL
• NHS-R community
  open source community of R users in NHS
• R course at Imperial Graduate School
  Online course
• R for Data Science
  Great textbook for learning R with principles of data science
• Introduction to Statistical Learning
  Text book for statistics (and a couple ML) topics
• The Elements of Statistical Learning (PDF)
  Foundational text for intersection of statistics and ML
  
  
• Crash Course Statistics
  Series of videos that straightforwardly explain statistics concepts
• NHS Python Community

Please see question "What resources are avaiable to help me with data analysis?" above.

For information on the available datasets visit our Information for researchers page.

If you are working with a fellow member of your research team who has the same access to data on the iCARE Environment as you, then you can collaborate in your analysis work. This could mean sitting at the same computer with the iCARE Environment open, or one of your sharing your screen during a video call. If however someone is present who does not have access to the same level of data as you, then this is not permitted and is considered a data breach. This includes people such as the project’s Clinical Sponsor.

Any outputs shared with people who do not have access to the data must follow the standard process for extracting files from the iCARE High Performance Environment. Please see the FAQ which outlines what can and cannot be removed from the Environment.

Information to be updated soon.

Please refer to the Metadata Catalogue on the iCARE website.

It is imperative that you acknowledge the iCARE team in all publications that utilised iCARE data and resources. This is outlined in the Terms of Use for iCARE Data signed by researchers. Please included the following text in your paper’s acknowledgements section: “The research was enabled by the Imperial Clinical Analytics Research and Evaluation (iCARE) environment and used the iCARE team and data resources.”

Please submit any questions regarding the data model to our email address.

Researchers should not remove any outputs whatsoever from the iCARE High Performance Environment; this is outlined in the Terms of User for iCARE Data which each researcher is required to sign prior to gaining access to the Environment. The platform has also been designed so as to not allow files to be removed. Should you require documents to be removed therefore, please contact us and indicate which files you would like extracted and where they are located. These documents will then be emailed to you directly.

Please note that there are restrictions on what can and cannot be removed from the iCARE High Performance Environment. All tables, figures and visualisations relating to iCARE data must be aggregated. Files containing individual patient-level will not be removed from the Environment. Similarly, any patient counts that are less than five must be replaced with a <5 in order to be extracted.

The iCARE Team is able to undertake data discovery work to try and locate routinely captured patient data that is not currently available in the iCARE Environment. No guarantees can be provided however on whether the data can be located, and how long this will take. Should the data discovery work required be significant, the project team will need to provide funding for this work.

If you suspect that a data breach has occurred, email the iCARE Team at Email: Imperial.dataAccessRequest@nhs.net immediately. Please provide details on what has happened.

The iCARE Environment allows for analysis to be conducted using R and Python.

Please refer to the section of the iCARE website relating to ongoing research projects.