Introduction to the Institute for Clinical Evaluative Sciences (ICES) and ICES expansion site, ICES@Western November 7, 2012 CPAH/RDC Statistics and Data Series at Western Amit Garg, Director ICES@Western Theresa Hands, Privacy Officer Salimah Shariff, Lead Epidemiologist Enhancing the effectiveness of health care for Ontarians through research 2 Outline • Who are we? • What does ICES do? • What does ICES@Western offer? • Where is ICES@Western located? • When will ICES@Western be operational? • How does ICES do it? Enhancing the effectiveness of health care for Ontarians through research 3 Who are we? ICES • Independent, non-profit organization • Conducts research that contributes to: effectiveness, quality, equity and efficiency of health care and health services in Ontario Enhancing the effectiveness of health care for Ontarians through research 4 Who are we? Network of Health Services Research Centres Opened 2012 Enhancing the effectiveness of health care for Ontarians through research 5 What does ICES do? ICES goals • • • Carry out population-based health services research that is relevant to clinical practice and health policy development Document province-wide patterns and trends in health care delivery Develop and share evidence to inform decisionmaking by policy makers, managers, clinicians, planners and consumers Enhancing the effectiveness of health care for Ontarians through research 6 What does ICES do? ICES goals • • Promote linkages among health services researchers and decision-makers Train researchers and promote a wider understanding of epidemiology and health services research Enhancing the effectiveness of health care for Ontarians through research 7 What does ICES do? Examples of ICES studies • Some of these you may have heard of…. Mortality among Patients Admitted to Hospitals on Weekends as Compared with Weekdays (NEJM, 2001) Cyclo-oxygenase-2 inhibitors versus non-selective nonsteroidal anti-inflammatory drugs and congestive heart failure outcomes in elderly patients: a population-based cohort study (Lancet, 2004) Unwalkable neighborhoods, poverty, and the risk of diabetes among recent immigrants to Canada compared with long-term residents (Diabetes Care, 2012) Enhancing the effectiveness of health care for Ontarians through research ICES Ontario, Canada linked database characteristics Cost: if had to collect data same data prospectively or through medical chart review very expensive. Question Do kidney donors compared to non donor controls of similar health have a higher risk of major cardiovascular events when followed for years after donation? Garg AX et al. BMJ 2012 We conducted a retrospective, population-based, matched cohort study that used large Ontario, Canada population-based healthcare databases Ontario Residents Restriction: Excluded any Ontario resident with a medical condition before their index date that could preclude donation: i.e. population kidney disease, DM, HTN, Cancer, Randomly assigned an index date to entire Ontario (using same distribution of index dates in donors)CVD, pulmonary dx, liver dx, SLE, chronic infections, HIV This provided on average of 11 years of medicalFrom records for baseline assessment, 9.6 million Ontarians this resulted in with 99% of people having at least two years ofexclusion baselineof data forofreview 85% adults (15% remained) Healthy Non-Donors Living Kidney Donors 1992-2009 - Manual review of all charts Match each donor to 10 non-donors Date of nephrectomy (controls) AgeNon-Donors (±2 years), Sex referred to asIndex index date Date health (±6 months) of similar as donors (start time ofIncome follow-up) and Residential Status Some Baseline Characteristics Living Kidney Donors Non-Donors (N = 2028) (N = 20,280) Age at Index Date, Years (IQR) 43 (34-50) 43 (34-50) Age Last Follow-Up, Years (IQR) 50 (42-58) 50 (42-58) 60% 60% 6.8 (0.5 to 18) 6.5 (0.1 to 18) 15,176 147,332 Women Follow-Up, years (range) Follow-up, total-person years Validated codes were used Major cardiovascular event (death censored) Good precision in the estimate No difference between the groups Donors: 1.3%; 1.7 events per 1000 person years Non-donors: 1.4%; 2.0 events per 1000 person years Hazard ratio 0.85, 95% CI 0.57 to 1.27 Population database strengths • Large population-based assessment • Databases allow for rigorous selection of non-donor controls • F/u up to 18 years, minimal loss to follow-up. (<6% emigrated) • CV events based on accurate and reliable codes (no recall bias) • Cost of this study $ vs. prospective study $$$$$$ Population database limitations • Association vs. causation. Unmeasured confounding • Some data not available: body mass index, blood pressure values, creatinine, cause of death 2) Hepatocytes Statin Metabolism CYP3A4 Metabolized Statins Atorvastatin, simvastatin, lovastatin 2% 40% 2% Systemic Availability 1) Enterocytes - Small Intestine 100% 40% Statin Metabolism + CYP3A4 inhibitor 2) Hepatocytes CYP3A4 Metabolized Statins + Clarithromycin / erythromycin Clarithromycin or Erythromycin 40% inhibits CYP3A4 40% Systemic Availability 80% 1) Enterocytes - Small Intestine Azithromycin a macrolide antibiotic used for similar indications does not100% inhibit CYP3A4 80% Population Based Study of Older statin users with new outpatient co-prescriptions Can’t ethically randomize Question: 30 severe days a patient to a potentially adverse interaction. Clarithro drug-drug / Hospitalization with Erythro rhabdomyolysis (n ~ 75,000) Hospitalization with AKI Azithro (data in subpopulation) Would be required to intervene All cause mortality to atcompared time knew about+worrisome AKI based on serum creatinine prescription. (n ~ 68,000) Patel A et al. (under review) 2012 Baseline Characteristics by co-prescription Clarithromycin n=72,591 Erythromycin n = 3,267 total n = 75,858 Azithromycin n = 68,478 74 (SD 6) 74 (SD 6) 40,130 (53%) 36,323 (53%) Cerebrovascular disease 3,189 (4%) 2,765 (4%) Peripheral vascular disease 2,101 (3%) 1,844 (3%) Coronary artery disease 39,908 (53%) 36,950 (54%) Congestive heart failure 12,652 (17%) 11,776 (17%) Systemic malignancy 21,875 (29%) 19,955 (29%) Demographics Age, years Women Co-morbidities Clarithromycin n=72,591 Erythromycin n = 3,267 total n = 75,858 Azithromycin n = 68,478 Atorvastatin 55,027 (73%) 50,111 (73%) Simvastatin 18,421 (24%) 16,369 (24%) 2,410 (3%) 1,998 (3%) High dose statin 30,296 (40%) 27,550 (40%) Low dose statin 45,562 (60%) 40,928 (60%) Oral hypoglycemic or insulin 20,367 (27%) 17,819 (26%) Beta-blockers 29,318 (39%) 27,008 (39%) 7,941 (11%) 7,206 (11%) 18,521 (24%) 16,982 (25%) 3,307 (4%) 2,992 (4%) Non-potassium sparing diuretics 26,901 (36%) 24,720 (36%) NSAIDs (excluding aspirin) 16,516 (22%) 14,797 (21%) ACE inhibitor or ARB 49,017 (65%) 44,323 (65%) Statin Characteristics Lovastatin Medication use in preceding year Verapamil or diltiazem Use of other calcium channel blockers Potassium sparing diuretics Clarithromycin n=2,334 Erythromycin n = 93 total n = 2,427 Azithromycin n = 1,488 serum creatinine, µmol/L 90 (76-108) 90 (76-108) eGFR mL/min/1.73 m2 66 (51-80) 65 (51-79) ≥ 90 mL/min/1.73m2 170 (7%) 109 (7%) 60-89 mL/min/1.73m2 1294 (53%) 782 (53%) 45-59 mL/min/1.73m2 564 (23%) 331 (22%) 30-44 mL/min/1.73m2 281 (11%) 199 (13%) 118 (5%) 67 (5%) Renal Function eGFR category < 30 mL/min/1.73m2 Results Number of Events (%) Clarithromycin/ Erythromycin n=75,858 Azithromycin n=68,478 Adjusted Relative Risk (95% CI) Rhabdomyolysis 24 (0.03%) 10 (0.01%) 2.17 (1.03 to 4.52) Acute kidney injury 347 (0.46%) 176 (0.26%) 1.78 (1.52 to 2.19) Mortality 529 (0.70%) 306 (0.45%) 1.56 (1.37 to 1.82) Number needed to harm (death) 1 in 400 Subpopulation with Lab Values Number of Events (%) Acute kidney injury (AKIN stage 1) Clarithromycin/ Erythromycin Azithromycin n=2,427 n=1,488 47 (1.94%) 10 (0.67%) Relative Risk (95% CI) 2.92 (1.47 to 5.79) Number needed to harm (AKI) 1 in 80 Patel A et al. (under review) 2012 Strengths First population based study looking at outcomes of this drug-drug interaction. Consistent with pharmacokinetic and case report data. Large number of patients. Near identical baseline characteristics table for both groups. Limitations Associations, therefore not causal. Unable to analyze by specific statin type. Analysis restricted to older adults. 25 How does ICES do it? Key to ICES research • • Ability to anonymously link population-based health information on an individual patient basis Linked data allows researchers to obtain a more comprehensive view of specific health care issues Enhancing the effectiveness of health care for Ontarians through research 26 Ms. Theresa Hands Privacy Officer, ICES@Western Enhancing the effectiveness of health care for Ontarians through research 27 How does ICES do it? How health data is generated • • • Patient enters hospital to receive care Person sees physician for a consult Elderly person lives in nursing home CONSENT • Given to the institution / caregiver only • Can be explicit or implicit • Use data for care, or for improving future care Enhancing the effectiveness of health care for Ontarians through research 28 How does ICES do it? How health data is generated • Patient enters clinical study • Person fills out survey EXPLICIT CONSENT • to use data for specified purpose Enhancing the effectiveness of health care for Ontarians through research 29 How does ICES do it? Ontario Privacy Law: PHIPA (2004) • Personal Health Information Protection Act • No use of Personal Health Information (PHI) without consent Enhancing the effectiveness of health care for Ontarians through research 30 How does ICES do it? So… how can we do our work? EXCEPTION: “Prescribed Entities” • 4 of these in Ontario Pediatric Oncology Group of Ontario (POGO) Cancer Care Ontario (CCO) Canadian Institute for Health Information (CIHI) Institute for Clinical Evaluative Sciences (ICES) Enhancing the effectiveness of health care for Ontarians through research 31 How does ICES do it? Privacy • • • ICES is a prescribed entity under PHIPA 2004 (s. 45[1] and O. Reg 329/04 section 18[1] ): Disclosure for planning and management of the health system Health information custodians may disclose personal health information (PHI) to ICES for the purposes of: Analysis and compiling statistical information related to managing the health system allocation of resources evaluation and monitoring planning for all or part of the health system Enhancing the effectiveness of health care for Ontarians through research 32 How does ICES do it? Prescribed Entities • • Have policies, practices and procedures in place to protect the privacy interests of the individuals and the capacity to make sure it is kept securely The Ontario Information and Privacy Commissioner (IPC) has reviewed and approved these policies, practices and procedures • The review / approval is renewed every three years • VERY big deal! Enhancing the effectiveness of health care for Ontarians through research 33 Dr. Salimah Shariff Lead Epidemiologist, ICES@Western Enhancing the effectiveness of health care for Ontarians through research 34 How does ICES do it? Key to ICES research • Ability to anonymously link population-based health information on an individual patient basis Enhancing the effectiveness of health care for Ontarians through research 35 How does ICES do it? Data sources – Patient-level data • Health Services Databases “transactions” of health care utilization Ontario Health Insurance Plan Claims (OHIP) Discharge Abstracts for Inpatient Hospitalizations (CIHI DAD) Same Day Surgery (CIHI SDS) National Ambulatory Care Reporting System (NACRS) Ontario Drug Benefit Claims (ODB) Enhancing the effectiveness of health care for Ontarians through research 36 How does ICES do it? Data sources – Patient-level data • Disease Cohorts/Registry Databases Received from partner organization Ontario Cancer Registry (OCR) Canadian Organ Replacement Register (CORR) ICES Derived Databases Ontario Diabetes Database (ODD) Linkage of delivering mothers to their newborns (MOMBABY) Enhancing the effectiveness of health care for Ontarians through research 37 How does ICES do it? Data sources – Patient-level data • Externally Linked Data Sources Chart reviews Electronic medical records (ex. Cerner) Surveys • Statistics Canada (Canadian Community Health Survey - CCHS) • Interviews, questionnaires Research datasets Enhancing the effectiveness of health care for Ontarians through research 38 How does ICES do it? Data sources – Patient-level data • Patient-level data sources can be linked to each other using unique patient identifiers ICES Key Numbers (IKN)= Encrypted OHIP # Enhancing the effectiveness of health care for Ontarians through research 39 How does ICES do it? Data sources – Supporting data • Population and Demographic Data Registered Persons Database files (RPDB) Population Estimates Canadian Census Profiles PCCF+ (StatsCan postal code conversion file) Citizenship and Immigration Canada Enhancing the effectiveness of health care for Ontarians through research 40 How does ICES do it? Data sources – Supporting data • Other Supporting Data Sources Care providers (all physicians in Ontario) Facilities and Institutions Management Information System (MIS; financial and statistical data for hospitals) Geographic conversion tables (LHIN, Census Subdivisions, Counties, Residence codes, etc) Coding tables (ICD9, ICD10, CCP, CCI, OHIP) Enhancing the effectiveness of health care for Ontarians through research 41 How does ICES do it? Linking datasets • All data sources can be linked at various levels IKN Health care facilities Facility Identifier Patientlevel Postal Code, LHIN Population & Geography Enhancing the effectiveness of health care for Ontarians through research RPDB Physician Identifier Care providers 42 What does ICES@Western offer? Process for intaking ICES studies 1. Develop a research question Do we have the data you need? 2. Identify ICES Scientist Only ICES Scientists can access ICES data for research Principal investigator or co-principal investigator must be an ICES scientist Must participate fully in the research project Responsible party within ICES Enhancing the effectiveness of health care for Ontarians through research 43 What does ICES@Western offer? Process for intaking ICES studies 3. Submit 1-page research proposal & Privacy Impact Assessment (PIA) Reviewed by a lead Scientist PIA used to ensure privacy requirements at ICES mandated by the “Personal Health Information Protection Act” (PHIPA) Ensures that project conforms to ICES policies and procedures Enhancing the effectiveness of health care for Ontarians through research 44 What does ICES@Western offer? Process for intaking ICES studies 4. Submit Project Activation Worksheet (PAW) Required in order to activate the project Provide estimate of resources and time for project (budget) 5. Request REB approval from Western Submit to Health Sciences REB, Delegated Review Enhancing the effectiveness of health care for Ontarians through research 45 What does ICES@Western offer? Process for intaking ICES studies 6. Develop Dataset Creation Plan (DCP) Study design Data sources and data elements Time-frames Inclusion/Exclusion criteria Variable definitions • • • Exposure Outcomes Other variables Analytic plan Output tables Enhancing the effectiveness of health care for Ontarians through research 46 What does ICES@Western offer? Process for intaking ICES studies 7. Submit project for execution at ICES@Western Reviewed by lead epidemiologist/analyst Entered into the queue Analyst assigned • • Only ICES staff and faculty have access to ICES data holdings Analysts are trained and have expertise in conducting health administrative database research Enhancing the effectiveness of health care for Ontarians through research 47 What does ICES@Western offer? Process for intaking ICES studies 8. Project commences Continuous communication with analyst & study team May require modifications to DCP 9. Results provided Aggregate level Results with fewer than 6 individuals are reported as “<=5” 10.Publish! Enhancing the effectiveness of health care for Ontarians through research 48 What does ICES@Western offer? Things to consider • Cost Cost recovery Grant support letter • Project must have grant (or other) money • Staff time is billed to investigator • Must be submitted at least 4 weeks prior to grant submission deadline • Time Projects vary in time depending on scope and resources required Enhancing the effectiveness of health care for Ontarians through research 49 What does ICES@Western offer? Services we can provide • Epidemiological • Biostatistical • Literature searches Study design/methodology, DCP advice Analytic design advice Administrative Grant support letters REB submission New data linkages Enhancing the effectiveness of health care for Ontarians through research 50 What does ICES@Western offer? Services we can provide • Support for Graduate Student Training Work closely with an analyst Have limited access to data Can perform own analyses Enhancing the effectiveness of health care for Ontarians through research 51 What does ICES@Western offer? Services we can provide • Faculty Scholars Program www.iceswestern.ca Open to all Western Faculty Training, mentorship, epidemiologic and analytic support to help Scholars develop and complete a minimum of one ICES study Guidance on grant writing and manuscript preparation Applications accepted until February 2013 for program commencement in September 2013 Enhancing the effectiveness of health care for Ontarians through research 52 When & Where ICES@Western • When will we be operational? • January, 2013 Process for new projects can start NOW Where are we located? ELL-218 (within Pharmacy hallway) Victoria Hospital, E-Tower, Lower Level Enhancing the effectiveness of health care for Ontarians through research 53 Contact Us ICES@Western • Amit Garg, Director ICES@Western • Theresa Hands, Privacy Officer • amit.garg@lhsc.on.ca 519-685-8502 theresa.hands@lhsc.on.ca 519-685-8500 x56045 Salimah Shariff, Lead Epidemiologist salimah.shariff@ices.on.ca 519-685-8500 x56555 Enhancing the effectiveness of health care for Ontarians through research 54 Questions? Enhancing the effectiveness of health care for Ontarians through research 54 of XX