Enhancing the effectiveness of health care

1
MOHLTC/Hospital Data Blitz Session Toronto April
10, 2006 Karey Iron
Enhancing the effectiveness of health care for
Ontarians through research
2
Objectives

• What /who is ICES?
• What does ICES do?
• Why is your work so important for ours?

3
What is ICES?

• ICES Institute for Clinical Evaluative Sciences
• Independent
• Non-profit
• Funded by Ontario Ministry of Health and
  Long-term Care core funding and peer-reviewed
  research grants
• Core activity
• To produce evidence to guide decision making in
  health care

4
G-Wing, Sunnybrook Campus
Enhancing the effectiveness of health care for
Ontarians through research
5
Who is ICES?

• Scientists (MD PhD Pharm U of T)
• Research Coordinators
• (project managers, epidemiologists)
• Biostatisticians
• Programmers
• Administrative assistants
• Knowledge transfer
• Library systems
• IT and Systems

Enhancing the effectiveness of health care for
Ontarians through research
6
Who are we?
Enhancing the effectiveness of health care for
Ontarians through research
7
What does ICES do?

• Use population-based health information to
  produce knowledge on a broad range of health care
  issues in areas such as
• cardiac care
• cancer care
• emergency services
• primary care
• prescription drug utilization
• population health
• health system performance
• Produce unbiased evidence to inform
• policy and practice

8
What does ICES do?
Atlases Investigative reports
9
HIPS
10
(No Transcript)
11
What does ICES do?
Peer reviewed articles
12
Trends in Deliveries, Prenatal care, obstetrical
complications in women with pre gestational
diabetes Feig D, Rassaz A Sykora K Hux J
Anderson G Diab Care 2006
13
Effects of socioeconomic status on access to
invasive cardiac procedures and on mortality
after acute myocardial infarction Alter DA
Naylor CD Austin P Tu JV. NEJM 1999
Survival post-AMI by SES
High income
Low income

14
What does ICES do?
Dissemination and education
15
How do we do it?

• We rely heavily upon administrative health data
  collected by hospitals, health care providers,
  and government
• Discharge Abstract Database
• National Ambulatory Care Reporting System
• Ontario Health Insurance Plan
• Ontario Drug Benefit Plan
• Registered Persons Database
• Corporate Provider Database
• Statistics Canada Census
• Others

16
How do we do it?

• and sometimes supplement these data with
  information gathered via...
• Chart reviews
• Surveys interviews, questionnaires
• Clinical registries
• Other

17

• but the core information
• we rely upon are the administrative health data
  produced by you every day!
• Why are these data so valuable?

18
The Strength is in the Linkage
19
Registered Persons - Demographics - Encrypted
Identifier Birth/Death Date Gender Area of
Residence
Drug Claims - Drug Utilization - Encrypted
Identifiers Prescription Date DIN Quantity
Encrypted Identifiers
CIHI DAD - Hospitalizations - Encrypted
Identifiers Admission/Discharge
Date Diagnoses Procedures
Physician Claims - Outpatient Visits - Encrypted
Identifiers Visit Date Diagnosis Physician
Specialty
20
Examples of the Power of Linkage

• RPDB provides demographic information
• OHIP provides information about health service
  utilization
• Census permits estimation of disease prevalence,
  rates of service use, and impact of SES
• CIHI permits assessment of disease and health
  service outcomes

21
Primary care visits (OHIP)
SES (Census)
Hospital admissions (DAD)
ED visits (NACRS)
AMI patients in Ontario
Survival to discharge (DAD)
Rx Drugs Post-discharge (ODB)
Post-discharge survival (RPDB)
ECG/lab data (Medical record review)
22
Example using administrative data for chronic
disease surveillance

• Chronic diseases have major impacts on patients
  and resources
• Survey data may underestimate these
• Can we do better with administrative data?

23
Identifying People with Diabetes (DM)

• Based on work from Manitoba, found persons with 2
  OHIP claims or 1 DAD/NACRS record indicating
  diabetes in past 2 yrs
• Compared to data from primary care charts
• Detect 86 of cases
• Mislabel less than 3 of cases

24
Number of Persons with DM in Ontario
25
DM Prevalence among Womenby Age and SES
26
DM Prevalence by County
27
Who Provides DM Care?
6.9
17.8
0.9
74.4
Fiscal 1998 - 2000
28
Independent Predictors of Acute Complications of
DM
Odds Ratios -0.5 1.0 1.5
2.0
No MD visits in prior year Residence
(rural/urban) Region (per region) Income (per ?
quintile) Sex (female vs. male) Comorbidity An
nual MD Visits Regular source of care Age
(per decade) Seen by specialist
29
Risk factors for Amputation amongthose with DM
30
Issues with DM Data

• Missing data
• Laboratory testing and results
• Services delivered by non-physician providers
• Non-OHIP physician care
• Timeliness
• Want to report current events rather than history
  data lag
• Coding Specificity
• Gestational diabetes vs pre-existing diabetes
• Side of amputation

31
Summary

• Population-based linked databases translate data
  gt evidence gt better decisions
• Overall, data quality is good......
• but, the full potential of these data cannot be
  realized without continuous improvement of data
  quality
• Hospital administrators, clinicians, and health
  record personnel all have roles in helping to
  ensure the highest quality data possible
• Continuous dialogue between creators and users of
  the data is key

32
www.ices.on.ca