Friday, 26 June 2015

ConnectingGTA: Finally a Shared EHR in Ontario

Ontario’s been struggling for years to get some form of Shared Electronic Health Record (EHR) off the ground.  The closest thing they’ve had so far was South-Western Ontario’s ClinicalConnect (not so much of a shared EHR as a hub into the various systems in the region) or the Integrated Assessment Repository (used mostly for Mental Health Assessments across the province).

ConnectingGTA is really the first major centralized Clinical Data Repository (CDR) that Ontario has built, thanks to some strong leadership from UHN and the ConnectingGTA Clinical Working Group.  I predict that the ConnectingGTA CDR will become the dominant EHR in the province that all other shared health records initiatives will end up integrating into.  This is a big deal.

Last week, I had my first meeting with the ConnectingGTA team.  Here are some things I learned:
  • The CDR contains all patient medical records from 17 sites dating back to May 2013: 27 Terabytes.
  • 2 million encounters from 17 sites are added to the CDR every week.
  • OLIS lab results are integrated at the field level, so graphs of vitals are available.
  • Most other data are stored as unstructured documents.
  • The main gap in the data is medication and primary care.
  • There is a project underway to integrate ODB data into the system and to import ClinicalConnect data (which should bring with it some primary care data).
  • In the initial pilot, of the 1200 pilot users, 40% logged in, which is higher than the average pilot program that would typically see 20% participation.
  • The system is onboarding thousands of new users every week.  Currently over 8000 users are signed up.
  • Main usage is in hospitals and CCACs.  It has revolutionized the transition of care from CCAC to Long-term care.
  • Most are logging in using federated access (i.e. they login to their source system and then click-through to ConnectingGTA without needing to login again.)  When they click through from a patient in their local system, they automatically get directed to that patient’s page in ConnectingGTA.
  • Some users are logging in directly through ConnectingGTA because the system-to-system click-through option can be slow.  In this case, they need to manually search for the patient by MRN, OHIP, or demographics.
  • The system currently uses UHN’s in-house client registry (list of patients.)  The ConnectingGTA team would like to use the provincial registry, but the current mainframe based one (that was a rudimentary extension of the old OHIP system) won’t meet ConnectingGTA’s requirements.  The ConnectingGTA team recommends that the province upgrade to a more modern Client Registry before switching ConnectingGTA over to the provincial client registry.
  • ConnectingGTA is on a fixed release schedule of 2 releases/year.
  • For communities outside of the GTA, they will only be granted access to ConnectingGTA if they first contribute data.  This is for 2 reasons:
    • It’s an incentive to get more data into the system.
    • If ConnectingGTA is introduced into an organization with none of their patient data in it, most patient searches will come up empty and clinicians would stop using it.
  • Currently there are no plans for “secondary use” of the data.  (E.g. analysis of outcome measurement or development of clinical guidelines.)
My main takeaway from the presentation is that the viewers would be far more useful if the data coming in was structured.  If a physician needs to review a patient, are they really going to sift through huge stacks of electronic documents?  That being said, perfection is the enemy of good, and having a stack of documents to read is better than having no documents at all; Physicians are used to working this way—this new system is just moves the stack of documents from a clipboard to a computer screen.  I just wonder if increasing the stack from 10 documents to 100 might dampen their interest in reading any of them…

The other thing that caught my eye was the comment about the Client Registry.  Having a shared understanding of what patient you’re talking about is essential to any EHR initiative in the province, whether they be drug systems, lab systems, online appointment booking or electronic referrals.  Ontario has taken far too long to get a functioning client registry.  So far, Hospital shared Diagnostic Imaging is the only provincial infrastructure that is using the provincial client registry.  That’s just embarrassing.

If I was running eHealth Ontario, I would have a poster on the wall with two numbers on it:
  • Number of projects waiting for access to the Client Registry
  • Average length of time to onboard a project onto the Client Registry (hint: this should be measured in days, not years)
And I’d bonus the execs on how low those two numbers are.

1 comment:

  1. Ken, thank you very much for the excellent post. I totally agree that perfection is the enemy of good and that it's better to have a non-perfect working system (with a plan for gradual improvements) than a perfect, but not working one. In regards to the Client Registry, my understanding is that the problem is not technical/performance related (it's likely got confused with performance limitations of Provider Registry, which should be resolved in the next release of PR/EMPI), but is rather related to privacy and policy: the ministry needs to approve clinical use of CR's administrative data for each onboarding system. I hope eHealth Ontario recognizes this issues and will act soon to eliminate all barriers for province-wide use of the Client registry.