It’s the second day of the minimally invasive surgery course at Monaldi Hospital and there are a score of Italian physicians speaking in addition to the main events – Dr. Henrik Hansen and Dr. Diego Gonzalez Rivas.
One of the surgeons addressing the group this morning is Dr. Andrea Droghetti, a thoracic surgeon from Carlo Poma Hospital in Mantova, Italy. Dr. Droghetti is here to present the latest information on the Italian VATS registry, Vatsgroup.it.
Now long-term readers know how we love a registry! We have talked a lot about the STS thoracic database and how it is woefully underutilized, we’ve talked to other surgeons who have been involved in creating their own national databases, and we even created our own.
As we discussed during a recent interview, data collection and publication are essential for research and advancement of the specialty – and that all starts with accurate data and statistics. But not all data collection tools are alike.
It is interesting, and encouraging to hear about the successful enrollment of 57 Italian facilities into a nationwide VATS registry to track VATS surgery and outcomes.
The database itself is pretty detailed and much more involved than the high altitude database or even STS. There are multiple risk stratification measures as well as quality of life indicators. The database is also designed to allow greater categorization – of pre-operative conditions, staging, procedures, and just about anything else you can think of.
Sounds like a great way to improve the quality of the data being used for research. After all, plenty of surgeons in Italy are participating – and as we know, getting surgeons to participate is always difficult. Even the STS database is lagging with just over 215 surgeons participating.
That’s awesome.. Now if only we could get more global participation!
Unfortunately, these kinds of large-scale projects never go off without a hitch – and during the presentation, we noted several potential pitfalls. One the major ones that Dr. Droghetti addressed was:
– Getting surgeons/ hospitals to participate
Out of 57 sites that are eligible to participate, only 44 are actually submitting data, and the data volumes have been measly – at just over 2 cases per day. (There is certainly more than two cases being performed every day.)
It also makes you wonder about the ‘randomness’ of the cases being entered. Maybe it’s one very diligent site entering cases everyday, or maybe it’s different sites entering their best outcomes – so the potential for data skewing seems to be there.
But since it seems like such a great project, Thoracics.org asked Dr. Droghetti to talk to us some more about this project, (translational issues during the conference made parts of the presentation unclear) and answer some additional questions. He was nice enough to talk to Thoracics.org for a few minutes.
From our own experience, we identified several other potential problems for the registry: so we posed these problems to Dr. Droghetti for his input.
Time consuming / repetitive entries for single patient
Data has to be entered on two occasions for the registry. The first submission takes approximately 30 minutes and the second – the post-surgical follow up – takes around ten minutes. The nice part about the project is that the patients actually participate in the follow-up evaluation and enter their own answers for the quality of life answers.
Now the QoL stuff is pretty unique to this registry, and the two entries per patient – allows for real-time time entry instead of retrospective review (which can get pretty skewed) so these are also strengths of the project. But..
After our own adventure with data collection as well as our experiences with the STS (cardiac) database, that this also immediately identifies this study as relying on 3rd party data entry. That’s because there is no surgeon under the sun that is going to spend that kind of time entering data when he could be seeing consults, performing surgery etc..
Third party data entry
is a dirty word in my book since it requires surgeons to rely on others to enter data about their outcomes. It’s also a negative because in many cases, the data entry is being done by a person who is more computer literate than medically literate. This means that they can’t always extrapolate data correctly from charts because they often don’t understand the data in the first place. This leads to unnecessary errors which skew data.
Dr. Droghetti and his team are addressing this issue, by appointing a specific “team member” but if that team member is someone specifically hired to enter your data (and not your anesthesiologist or other invested person) – then it’s no different from the third-party data entry systems we’ve seen before with STS (so expect similar problems). Computerized data entry tends to be tedious – and that might also be leading to the low participation rates we are seeing. With the amount of data to be entered, 30 minutes of drop down boxes might actually translate to more than an hour (just take a look at the cardiology PCI registry).
Hopefully these issues won’t impede Dr. Droghetti and his colleagues in their efforts. We wish them luck and look forward to seeing more publications based on this data.