Does it still apply that if an institution submits say 15,000 records to MPOG, any research requests from that institution should be limited to a dataset comprising of 30,000 records from MPOG database (15,000 from the institution + 15,000 from other sites)? In other words, available dataset is limited to whatever your home institution provided to MPOG + a matching number from other sites.
There is no matching case counts required. Once you have submitted your 10,000 cases, you have equal access to the MPOG dataset as another site that has contributed 500,000 cases. This is to incentivize smaller, community hospitals to join. We think that diversity of care pattern and academic interests is the key. The 10,000 minimum was created to avoid grossly unfair data access.
What if a parent institution is not submitting a certain type of data (say for example the institutions is not submitting outcome data), can a researcher from this institution request data for study that requires outcome data?
An organization can only request data that they are submitting. For example, if you are not submitting outcomes or lab values, you do not have access to outcomes or lab values in the PCRC process. This is done for two reasons: a) to incentivize hospitals to submit as many different types of data as possible and b) not having your data submitted makes it less likely that you understand the limitations (and opportunities) of a data source. If you haven’t gone through the process of transforming your data into the MPOG format for a given data type, the science will be more suspect.
Who makes the Perioperative Clinical Reserch Committe (PCRC)?
The PCRC is composed of each contributing center’s PI (or designate). We invite centers that are in process of but not yet fully contributing. However, the vote for a proposal is drawn from the centers that are contributing data.
Who exactly extracts data once a research project is approved by PCRC, a database programmer from the parent institution? Is there a time frame to complete this?
MPOG has programmers centrally located that extract the data. The MPOG programmer works with the PI to transform the proposal into a detailed data extraction specification. The time frame depends upon two factors a) the PI’s ability to respond to the programmer’s inquiries and b) other tasks on the MPOG programmer’s list. I would say that from proposal presentation to data in hand is typically 1 – 2 months. This includes the PI submitting a revised proposal to PCRC (for email review) and the programmer making the spec and PI responding. This was taking much longer earlier (3-6 months), creating much dissatisfaction. A centralized query activity allows us to make sure each project benefits from the MPOG programmers who know the idiosyncrasies of each site’s data.
What if the initial data extraction proved that additional data is needed to answer the research question. Should a modification be submitted to PCRC? Will it still need 60 days for approval?
If it is a minor, incidental additional data need, a new PCRC submission is not needed. If the initial data pull changes the hypothesis or analytical approach, a new PCRC submission is recommended.