Information for Researchers
The QResearch Advisory Group draws up the criteria for access that are applied by the QResearch Scientific Committee.
Researchers wishing to use the data should build in plenty of time into their schedule since it takes time to process applications and prepare data sets. Researchers will need to do as much work to help define their search criteria as possible and we are developing web tools to help with this. Researchers will, however, need to enlist clinical GP support in their institution to help define case definition for their searches.
QResearch releases data to academics employed by UK universities to undertake bona fide research. We assess eligibility of applications using the following questions.
- Did you have the original idea for this research project?
- To your knowledge is this work original and capable of publication as original research in a peer-reviewed journal?
- Are you free to undertake this study and publish its findings without needing to clear it with the funding source or any other organisation?
- Do you agree to acknowledge the source of QResearch data in any publication, paper, report or software/tool?
- Do you agree NOT to attempt to identify patient(s) or practice(s)?
- Do you undertake to provide a copy of the final report of the project and copies of any publications within one year of the project completion?
- Do you agree NOT to release the data to any third party including the funder, sponsor or other such body?
- Do you agree not to use the data for any other project except that which is expressly described in your protocol?
- Do you undertake to check the data you are given within a month of receipt and report back any problems within that time?
- Do you have a statistician on the project team who has contributed to the design of the study and will advise on the analysis?
- Do you agree to have a project summary on the QResearch website once the project starts?
An answer of 'no' to any of the questions is likely to invalidate the application.
If you are an academic employed by a UK university, please read the following information and also look at the publications page for examples of studies which have been undertaken.
How does it work?
Researchers originate a research question or hypothesis; write an outline protocol; and contact QResearch to discuss the feasibility of undertaking the study. If the study is feasible, QResearch will give a broad estimate of the costs of providing the data and will provide a letter to accompany any application for funding. The researcher then secures the necessary funding and completes the QResearch application form, including a detailed protocol and data specification. This application is sent for scientific review and feedback is given to the researcher. The researcher makes any necessary modifications to the protocol and approval is obtained, the researcher is given a timescale for the data extraction. Once the researcher has the data, they have to approve it within one month of receipt.
How much does it cost?
QResearch is not-for-profit and is entirely self-funding so needs to cover the costs associated with providing a research service. We will estimate the cost of each data extraction plus VAT. The estimate will take account of the complexity of the data required and the time taken to extract it. If the specification becomes more complex, then the cost will be uplifted. QResearch will submit an invoice to the Principal Investigator for the work concerned. This will be paid in advance of the data being extracted.
Who is eligible to apply for research access?
QResearch was established specifically to provide access for research projects by academics employed by UK universities. A license agreement will be between QResearch and the employing University. At least one member of the research team must be a medically qualified academic registered with the General Medical Council who signs the guarantee.
What types of studies are most suitable?
We will provide extracts of data on samples of patients with sufficient data to be able to test the hypothesis. The sample size will depend on the sample size calculation. We do not license the use of the whole database. Any research analyses which might involve the whole database or which need a practice level analysis need to be done on site and are subject to QResearch having adequate capacity.
The following studies are most suitable:
- case control studies designed to examine risk factors for onset of disease
- cross sectional surveys
- cohort studies
- sample size calculations
Are there studies which can't be done on QResearch?
Yes, there are some types of studies which are not suitable for QResearch. The QResearch team will let researchers know at an early stage if a study is not suitable. In particular, QResearch cannot undertake studies which involve:
- any intervention on practices or patients
- any need to contact the practices or patients since this is impossible
- any changes to the fundamental principles underpinning QResearch
- data which is not already held by QResearch.
Are there any data items which are not released?
In order to safeguard the identity of the practices, we do not release any practice level data or any individual Townsend Score. We also make special consideration of data items relating to patients with very rare or unusual disease. The linked hospital data (HES), cause of death supplied by the Office for National Statistics and cancer registration data supplied by Public Health England are only currently available for analysis on site at the University of Oxford as this is the license condition stipulated by the data suppliers. The HSCIC retains copyright© of HES data and it is re-used with permission of the HSCIC (all rights reserved). If QResearch is not a suitable data source for a project, we will endeavour to suggest other more appropriate data sources where possible.
What is the maximum number of patients that will be released in a dataset?
The number of patients in a sample depends on the sample size calculation.
What happens if a researcher contravenes the conditions of using the database?
Attempts to break the confidentiality of the practices or patients or clear evidence of research misconduct will be treated extremely seriously by QResearch and, if proven, will result in a bar on further access to data for the research team and their university. It will also lead to referral to the General Medical Council which is why one of the core teams needs to be registered with the GMC.
What if a pharmaceutical company wants to access data for research?
QResearch does not provide access to data for pharmaceutical companies except where there is a clearly defined concern regarding the safety of a drug. In these circumstances the pharmaceutical company will be referred to a third party appointed by QResearch to handle such requests. The normal governance procedures will be followed so that the analysis independent and published.
Does QResearch mind where the funding for my study has come from?
We ask you to declare the source of funding and we ask you to guarantee all the conditions above. All applications are reviewed by the Scientific Board and the MREC and if there are concerns then we will discuss it with you. If a researcher wishes to apply for funding for an application which has already been approved, and the grant application is based on, or contains reference to QResearch data, QResearch must be informed by email about the application before the application is submitted.
Duplication of academic studies
QResearch will inform prospective researchers if their project [or one very similar] is already being undertaken by another team in order to avoid duplication whilst maintaining confidentiality for both teams. We do not supply the same or very similar data to different teams to answer very similar questions. We also hope this will avoid disappointment and wasted effort by one team if the other team publishes first. However, if the second team can come up with good reasons why the analyses need to be performed again, then the QResearch Scientific Board will be asked to consider the best way forward.
QResearch undertakes research and analysis for a variety of organisations covering a wide scope of questions including both research and audit. Nothing in any agreement with a third party will prevent QResearch from supplying data and services to any other party at any time. At all times copyright in core material from the QResearch database licensed by any client remains the absolute property of QResearch.
What happens to the results of research projects?
We expect research to be published in peer-reviewed journals. Every form of publication arising from the work utilising the QResearch database shall acknowledge it as the source of the raw data in keeping with the conditions of access to the database. Publications should also refer to the version of the database and the date of last data download for the data used in the study.
What criteria are used for the Scientific review?
The Scientific Board judge applications against the following criteria
- Is there a clear research question or hypothesis which is likely to lead to generalisable findings capable of publication in a peer-reviewed medical journal?
- Are the researchers likely to be able to conduct the study and its analysis?
- Is QResearch the appropriate database to be used to conduct the research?
- Is the methodology appropriate to answer the question (including the possibility of bias and confounding)?
- Are there any potential risks to the ethical position of QResearch in undertaking this research (including the potential identification of patients or practices)?
- Do you have any conflict of interest in reviewing this application? If so, please give brief details?
- Overall do you think QResearch should approve this research project?
Why do we need to secure funding first?
We get many of enquiries from academics who then don’t proceed to undertake the project. This can be for a variety of reasons including a failure to identify the necessary funds. Our Scientific Board is unpaid and so we only ask them to review projects which are almost certainly going ahead. We ask the Scientific Board to review them before the MREC to ensure that the protocol which then goes to MREC doesn’t need to be amended and hence re-submitted.
I just want all the data for the patients of interest. Is this OK?
No. We expect researchers to have a specific research question/hypothesis that requires specific data. We do not license the whole database or give out entire medical records. The maximum number of patients which can be supplied is 100,000 and will usually be less than this.
What do I need to do if my research question changes?
The approval you gain is on the basis of the protocol submitted to the Scientific Board. If you would like to answer another research question which is not in your original protocol then you must re-submit it as a new project for peer review. An additional charge is likely since data are licensed on a per study basis.
What happens if my research team changes or I move institutions?
QResearch data may only be accessed by individuals and organisations and for the purposes of the research which has been approved. You may not transfer, distribute, share, sell or sublicense the data to any third party. Specifically, you may not transfer the data outside the UK. Should there be any changes in the individuals or organisations undertaking the research, you must notify QResearch so that an amended data agreement can be negotiated.
What if I don’t want to do the research but I want to know the answer?
QResearch may also do commissioned research if appropriate. This can be done in collaboration or as a service for which QResearch will give a formal quote. QResearch has a standard draft contract for this type of work which is available on request.
Is there an appeal process if my application to QResearch is not approved?
Reasons for refusal for an application include failure to comply with the essential criteria noted above or because the application fails to pass the scientific review process. It could also include failure by the applicants to produce academic papers from previous studies where QResearch data has been supplied. Should an application be rejected, the applicant is welcome to make the appropriate changes and submit a new application which would be expected to be substantially different to the original one. A covering letter documenting the reason for the appeal, what reasons had been given by QResearch for the rejection, and what actions and amendments have been undertaken should be included. A charge for processing the application may be raised to cover the cost of administering the appeal.
What happens if there is a mistake in the data?
QResearch will take every precaution to ensure that the data requested match the exact specification which has been agreed and signed off by the principal investigator. The data extraction process will be subject to routine quality assurance procedures. It is the principal investigator’s responsibility to confirm receipt of the data and check the data and notify QResearch of any errors or omissions within 28 days. In the absence of any such notification we shall consider that the data are correct. If there is a mistake in the data, then the principal investigator should contact QResearch who will investigate the problem.
If there is a mistake in the data extraction caused by QResearch error, then the principal investigator will be provided with a corrected dataset within 28 days or as soon as is reasonably possible. QResearch cannot be liable for the cost of any staff time spent working on the data.
If the mistake is due to an error in the data specification drawn up by the researchers (eg the researchers have selected the wrong Read codes), then QResearch will endeavour to help by re-running the data extraction routine with the new codes within the following 3 months. There will be an additional cost for this.
Publication of results
The results of all studies using QResearch data must be published and the results must be publically available to maximise the public benefit but also to reduce the risk of publication bias. The underlying individual level data for the paper must not be shared under any circumstances. Requests to access the individual level data must be directed to QResearch who will consider the request as a new application. Researchers must consider the risk of identifying individuals in their analysis prior to publication and must not publish small numbers (which are counts 1-5). A higher level of aggregation should be considered when analyses produce less than 5 cases.
How long do I have to undertake my research?
We will write to you after two years to invite you to submit for an extension to continue to use the data or to ask you to destroy the data if the project has finished and provide a report on the study.