○ Nirav Shah (MPOG): Great question. It would deidentified by both patient and
location perspective. Similar to how we do this for a research proposal- using an
anonymized institution ID.
■ Dan Kenron (OHSU) - deidentified refers both to PHI and site/location, correct?
○ Nirav Shah (MPOG): Correct (addressed in response to Dr. Esmail’s question)
■ Shafeena Nurani (Corewell East) – I would support It
■ Josh Goldblatt (Henry Ford Health, Jackson) - QI is typically an internal process. Analysis is
typically pre-post intervention. I can't picture a case where we would need data from other
sites. Can you give an example?
○ Nirav Shah (MPOG) - Site example was interested in looking at compartment
syndrome and if the incidence at their site was like other MPOG sites. Similarly, this
would allow sites to analyze data for a potential new QI measure and the feasibility
of this.
■ Silis Jiang (Weill Cornell): Any thoughts on requiring an IRB non-human subjects
determination?
○ Nirav Shah (MPOG): As most QI work is usually exempt by IRB standards, I think this
would fall into that category and also be exempt but would be governed by the
Quality Committee and would seek IRB approval for studies if there was human
subject involvement beyond our current performance measurement.
▪ Dan Biggs (University of Oklahoma) - If you identify an event that is rare enough, would that
allow us to identify the place?
▪ Nirav Shah (MPOG) - One of the ways we would want this approved is through quality
committee to discuss and manage that risk
▪ Kathleen Collins (Trinity Health – Livonia) - This might allow sites to check the frequency of
supposed/assumed rare cases or outliers. Some might be more frequent than we expect
▪ Tariq Esmail (University Health Network- Toronto) - On the research side, is there a
minimum case number that needs to be achieved to be revealed?
o Nirav Shah (MPOG): If you look at this in Data Direct at the cohort level, we have a
minimum threshold of cases to show up to reinforce the fact that there may be
some variation in quality
o Sachin Kheterpal (MPOG): This group is going to guide what that threshold is
o Tariq Esmail (University Health Network- Toronto): It sounds like this discussion is
just to determine if we should move forward and not the specific definition of
process but if one site has a complete outlier, we would need a process in place to
determine what to do with these types of circumstances to notify the site to let
them know. If the site has a rare event and is unaware of it, the Coordinating Center
would need to reach out to the site.
o Nirav Shah (MPOG): Yes correct. At the coordinating center we would want to know
we have approval from the quality committee to identify and reach out to the site
o Sachin Kheterpal (MPOG): What we’re discussing today is whether or not we should
allow QI project proposals. The specifics are very helpful to address any nuances.
▪ Shreyas Bhavsar (MD Anderson) – Thank you for bringing this forward. We originally were
looking at our local site data for compartment syndrome and with limited number of cases
were interested in the incidence of this complication among all MPOG sites.
▪ Kim Finch (Henry Ford Health System) - Is compartment syndrome something that would be
identified MPOG in the data? It seems like healthcare risk reporting system is where that