Transcript Health data in Ontario

Health data
in Ontario
Susan Bondy,
U. of Toronto Dalla Lana School of Public Health
Presented at:
Health Over the Life Course , Pre-conference Workshop
University of Western Ontario, October 14, 2009
[email protected]
Data sources
• Health surveys,
– Federal, Provincial, sub-provincial
•
•
•
•
Vital statistics data
National hospitalization data (CIHI)
Provincial health system data
Special disease registries, etc.
Health Surveys
• Ontario Health Surveys
– Custom in 1990; NPHS/CCHS buy-ins
• Rapid Risk Factor Surveillance System
(RRFSS)
• Ongoing thematic surveys, e.g.,
– CAMH OSDUHS (school survey since 1977)
and Adult “Monitor” surveys
Recipient data and linkage
• Registered Persons Data Base (RPDB)
– ‘accounts-level’ records for Ontario Health
Insurance Plan beneficiaries
– OHIN linkable to services funded by Province
– Not a registry of population, but of accounts
• E.g., death clearance is not aggressive
• Some research centres have created cleaner
versions
Hospital data
• All Ontario hospitals participate in CIHI
databases
– Emergency care, rehab. since 2000-2003
– Mental health facilities ~2005
• CCAC (home care) data system ~2005
• High quality data with patient, disease,
and care elements
Drug data
• Ontario Drug Benefits Plan:
– Residents over 65
• Fact and quantity in data
• Not dose, co-morbidity (or, necessarily, indication)
– If dispensed in hospital, or special cancer
drugs program
• Paid for; not necessarily in data
– Prescription drugs for <65 year olds
• Need-based provision (“Trillium Program”)
• For individual-level data, rely on self-report surveys
Drug data
• Some special drugs (tracked via…)
– Hospital-based dispensaries
– Government-controlled access
– Service fees for drug administration
– Clinical care electronic data
– All these examples apply to medical oncology
– Gaps for other patient groups
Ambulatory care / services
• Procedures and maneuvers in hospitals
observed via CIHI data
– Procedure codes
– Diagnostic data in same complex record
• Claims (billings) to OHIP for services by
registered providers (physicians and
others, e.g., physiotherapists)
“OHIP” data (i.e., claims data)
• Limited data on patient
– One diagnosis code (variation on ICD)
• Many opportunities for misclassification error
• Very little info. on context or intent
• What done, without why.
• Procedure codes of interest
– May be highly informative
• Specific to disease, purpose and provider
– May not exist as desired
• E.g., Pap. for screening (part of periodic exam; separate
billing only for diagnostic test)
– May be under-utilized
“OHIP” data (claims data)
• Increasing number and size of non-fee-forservices pockets
• Shadow-billings system supposed to
capture procedures
• Preventive services may be provided
(tracked) separately
– E.g., Provincial cancer screening programs,
other preventive programs (flu shot)
Special disease and treatment
registries (just a few examples)
•
•
•
•
•
Ontario Cancer Registry (OCR)
Ontario Familial Colon Cancer Registry
Ontario Trauma Registry
Ontario Diabetes Registry
Systemic Lupus International Collaborating
Clinics (SLICC)
• Ontario Cardiac Rehabilitation Registry (OCRR)
• ....
Acts of sharing
• Health Protection and Promotion Act, 1990
• Freedom of Info. and Protection of Privacy Act, 1990
• Personal Health Information Protection Act, 2004
– Health custodians in regulations:
•
•
•
•
1. Cancer Care Ontario.
2. Canadian Institute for Health Information.
3. Institute for Clinical Evaluative Sciences.
4. Pediatric Oncology Group of Ontario
• Health System Improvement Act, 2007
– Creates Ontario Agency for Health Protection and Promotion
Current state of access and
sharing
• Data access is possible
• By (and in partnership with)
• Recognized health custodians
• Partnership with Ministries of Health
• No truly open data library or warehouse
• Access is conditional on infrastructure, some
extension of access to researchers (e.g., ICESQueens and proposed elsewhere)
• Access improved in ~5 years,
• especially inside the system)
Partnering with data custodians
• Usually requires a Co-Investigator (or PI)
inside the custodian agency
• Understand that these are academics in
competitive settings, with restricted time for
plethora of requests
• Instantaneous partnerships have happened (don’t
always)
• Highly beneficial to have Government
interest
• A challenge for bureaucrats too (stretched;
regardless of intentions)
Selected external links
•
•
•
•
•
•
•
www.sph.utoronto.ca
www.ices.on.ca
www.cancercare.ca
www.rrfss.ca
www.camh.net
www.apheo.on.ca
www.chass.utoronto.ca/datalib/