The biggest problem we had trying to implement an OPPE was sourcing data. We had lots of different "metrics" from Excel files on 1 persons computer, cocktail napkins on another persons desk, Access database on a third persons desk. None of which could be counted on to be updated at any given point of time for synchronization.
Next problem was getting everyone to agree on the "metrics". Meaning does this count in the numerator/denominator?
Last problem was the "we need that to be excluded." That forced the system to have multiple people meet before any numbers could be "counted."
Hopefully my problems were unique
I think that Dalton summed up the issues most folks run into when diving into OPPE. One more important driver you'll need is executive backing. You'll need executives/leaders that are willing and able to wield their power to drive home the definition of metrics. If you have this, you'll be able to get through that part of the process in a much more time-efficient manner.
Once you have the metrics down, the remainder of the work is *simpler* in that it's usually easier to connect to and pull in data than it is to get agreement on metrics
Hope this helps, and don't hesitate to reach out if you want to talk further on this - more than happy to help.
Here's one database issue:
The code below shows all of the surgeons associated with a surgical case where there was a laceration or puncture comorbidity. But we do not know of a way to connect this to the surgeon.
SELECT DISTINCT TOP 1000 adds.SourceID, adds.VisitID, adds.AbstractID, adds.ComplicationComorbidity,
adds.Diagnosis, adds.DiagnosisName, aos.Name AS Surgeon
FROM livendb.dbo.AbsDrgDiagnoses adds
JOIN livendb.dbo.AbsOperations ao ON ao.SourceID = adds.SourceID
AND ao.AbstractID = adds.AbstractID
JOIN livendb.dbo.AbsOperationSurgeons aos ON aos.SourceID = ao.SourceID
AND aos.AbstractID = ao.AbstractID
AND aos.OperationSeqID = ao.OperationSeqID
WHERE ComplicationComorbidity IS NOT NULL
AND ( lower (DiagnosisName) LIKE '%laceration%'
OR lower (DiagnosisName) LIKE '%puncture%')
What metrics are you trying to monitor and/or what questions are you trying to answer? I know that different organizations are looking to answer different questions, so it would be helpful to know what your angle is here.
I have done some work in this area, but am not sure if my solution would be helpful to you and your situation. We have pulled together (joined) hospital and physician practice system datasets to allow for better system-wide reporting and analysis.
We're open to new metrics, as the ones we have been using have been limited because we didn't have Qlikview. So I'd be interested in what you are measuring.
But at this point we are having trouble attributing things to specific physicians. e.g.:
- How many surgeries has a provider performed? - right now we're getting surgeries attributed to pathologists and other non surgeons, and surgeons doing only 2 cases in a year when ancient Meditech reports show they've done 220
- If there was an accidental laceration during the surgery, who did it? Right now we can associate providers with surgeries and lacerations with surgeries but not providers with lacerations.
We've asked Meditech for help but they are very slow to respond (weeks, months).
Any help you could provide would be greatly appreciated!
Currently, our main clinical EHR includes a surgical module, so I'm not sure if this will be very helpful to you... We do have a fair amount of surgery-specific information that we maintain, but have focused on robotic surgery metrics in order to evaluate the utilization of existing surgical robots, as well as project the need for additional robots in the future. Some of our metrics include robotic cases by surgeon and procedure, as well as length-of-stay for those procedures and where our volume is coming from (patient zip code). We're also trying to compare non-robotic procedures to their robotic equivalents and although we have made some progress in some areas, this has proven very difficult to due a lack of understanding on how to equate robot/non-robot procedures for many types of surgeries (hernias, colon procedures, etc.). We have not delved into accidental lacerations and tracing those back to specific surgeons.
In terms of linking physician practice and hospital system EHR data sets- we link them within a QlikView application at the patient level, based on SSN, CPI, full name and/or a combination of name/address. That allows us to see some downstream results of patients that visited our physician offices (although a number of assumptions are made in order to do this).
QlikView has been an instrumental part of our decision-making process, as it has allowed us to bring together several disparate data sources for meaningful reporting and process improvement/BPM (business process management).
Does this help at all?
Hi I just joined the Qlik community and was researching something else for a work project but happened into this group because my background is in Healthcare quality and regulatory compliance.
I have built OPPE and Performance Improvement (compliance and audit) systems from the ground up in ambulatory settings. The most recent being a few years back, that was deployed across 5 hospitals in a large system.
The problem as others have mentioned is having a single source of agreed upon measures and dimensions, you will find they will diverge and branch over time in each setting (e.g. Hospital A vs Hospital B) and so you must account for that in the scalability and flexibility of the design. In this sense it is good to have Quality execs driving as well as a steering committee for consensus.
As to Qlik I am still very new to it but getting up to speed very rapidly.
For reporting OPPE you need 1 repository, you should be able to do everything you need in MS Access if budget is a concern, front end would support reporting as well.
Qlik has plenty of functionality around connecting to data sources, ODBC, MS SQL etc. But like anything, specifically quality reporting you need to ensure you avoid GIGO scenario's. If you have ever been caught on the other end of a meeting with Physicians and the data is faulty you will know why this is important.
There are a few software providers that have OPPE/Quality modules that are really good, The Advisory Boards, Crimson comes to mind, they are fantastic to work with I have PM'd 2 hospital implementations with them