Project overview
Background
Researchers often use anonymised GP records to develop tools that can help detect and predict diseases. However, these tools do not always perform accurately due to flaws in GP records. For example, GP records may not include information from all groups of people living in the UK. GP records may also be more likely to have missing or incorrect information for certain people, such as groups who don’t visit the GP regularly. Any tool developed using these records could result in inadequate care for these groups. A better understanding of what kind of information is missing from GP records and ways to handle this problem would help researchers develop better, fairer tools. This is particularly important as GP records are likely to be used to develop more tools in the future.
Aim
We want to improve understanding of missing information in GP records and ways to handle this.
Methods
We will search databases of published studies that have looked at and reported missing information in GP records and ways to handle this. We will search specific medical and informatics journals for additional studies. We will look for reports and relevant information from media and other non-academic resources using Google searches. Using these studies, we will describe what kind of information is missing in GP records and for whom. We will also examine how existing research using GP records has handled missing information. We will use our findings to produce an initial draft of a checklist that will consider and handle missing data.
Patient and Public Involvement (PPI)
Public contributors felt this was an important and timely topic. They discussed the need to build patient trust in the use of GP records for research and better transparency regarding potential harm. Public contributors noted it will be important to share our findings with members of the public, though this will need to be shared in a way that is meaningful to them. During the study, we will meet with public contributors to gain their input on our study plan, what our study findings mean to them, and how we can share our findings.
Impact
This study will lead to a future study to develop our checklist so that it can be used to help researchers who are developing tools from GP records consider and handle missing information in the records they are using. This will allow researchers to develop tools that more accurately identify and predict risk and outcomes for all groups of people.
Sharing our findings
We will share our findings in a scientific journal and relevant conferences. We will share our findings through social media and present them to members of the public, health professionals, researchers and local communities.
Researchers often use anonymised GP records to develop tools that can help detect and predict diseases. However, these tools do not always perform accurately due to flaws in GP records. For example, GP records may not include information from all groups of people living in the UK. GP records may also be more likely to have missing or incorrect information for certain people, such as groups who don’t visit the GP regularly. Any tool developed using these records could result in inadequate care for these groups. A better understanding of what kind of information is missing from GP records and ways to handle this problem would help researchers develop better, fairer tools. This is particularly important as GP records are likely to be used to develop more tools in the future.
Aim
We want to improve understanding of missing information in GP records and ways to handle this.
Methods
We will search databases of published studies that have looked at and reported missing information in GP records and ways to handle this. We will search specific medical and informatics journals for additional studies. We will look for reports and relevant information from media and other non-academic resources using Google searches. Using these studies, we will describe what kind of information is missing in GP records and for whom. We will also examine how existing research using GP records has handled missing information. We will use our findings to produce an initial draft of a checklist that will consider and handle missing data.
Patient and Public Involvement (PPI)
Public contributors felt this was an important and timely topic. They discussed the need to build patient trust in the use of GP records for research and better transparency regarding potential harm. Public contributors noted it will be important to share our findings with members of the public, though this will need to be shared in a way that is meaningful to them. During the study, we will meet with public contributors to gain their input on our study plan, what our study findings mean to them, and how we can share our findings.
Impact
This study will lead to a future study to develop our checklist so that it can be used to help researchers who are developing tools from GP records consider and handle missing information in the records they are using. This will allow researchers to develop tools that more accurately identify and predict risk and outcomes for all groups of people.
Sharing our findings
We will share our findings in a scientific journal and relevant conferences. We will share our findings through social media and present them to members of the public, health professionals, researchers and local communities.
