International Comparisons: Adding Value to the Canadian Health Care System

Transcription

1 International Comparisons: Adding Value to the Canadian Health Care System Panel presented at: The 9th Annual CAHSPR Conference Montréal, May 29,

2 Introduction and Opening Remarks Jeremy Veillard Vice-President, Research and Analysis CIHI 2

3 Panelists Chris Kuchiak, CIHI Heather Bryant, Canadian Partnership Against Cancer Jean-Frédéric Levesque, Institut national de santé publique du Québec G. Ross Baker, Institute of Health Policy Management and Evaluation, University of Toronto Health System Expenditures Cancer Systems Primary Health Care Patient Safety 3

4 Canadian Institute for Health Information International Comparisons: Adding Value to the Canadian Health Care System Comparing Health Expenditures Across Countries May 29, 2012 CAHSPR Panel Session Christopher Kuchciak, Manager, Health Expenditures 4

5 How Does Canada Compare? Total Health Expenditure as a Percentage of GDP, 29 Selected Countries, 2009 United States Netherlands France Germany Denmark CANADA Switzerland Austria Belgium New Zealand Portugal(*) Sweden United Kingdom Iceland Norway Spain Slovenia Finland Slovak Republic Australia(*) Japan(*) Czech Republic Luxembourg Hungary Poland Estonia Korea Mexico Turkey(*) Note * Data for Source OECD Health Data 2011, June edition. 12.0% 11.8% 11.6% 11.5% 11.4% 11.4% 11.0% 10.9% 10.3% 10.1% 10.0% 9.8% 9.7% 9.6% 9.5% 9.3% 9.2% 9.1% 8.7% 8.5% 8.2% 7.8% 7.4% 7.4% 7.0% 6.9% 6.4% 6.1% Average: 9.7% 0% 5% 10% 15% 17.4% 5

6 How To Facilitate Comparability Across Countries? National Health Accounts - a tool for describing, summarizing and analyzing expenditures of a country (Canada= National Health Expenditures (NHEX) database) Limitation: wide variations across countries in definitions and institutional settings OECD System of Health Accounts (SHA) Common framework to improve international comparability Distinguishes between functions (types of services) and providers. Facilitates comparability irrespective of the differences in health systems SHA version 1.0 published in 2000 by the OECD was the first manual to provide a framework for standard reporting of expenditure on health Canada an early adopter of SHA 1.0 6

7 How Comparable Are Health Expenditure Aggregates Across 34 Countries? OECD Health Data 2011 Fairly Comparable: 29 Countries Follow SHA manual Less Comparable: locally produced health accounts Even Less Comparable: national accounts data Australia Korea Turkey Ireland Chile Austria Luxembourg United Kingdom Israel Belgium Mexico United States Italy CANADA Netherlands Greece Czech Rep. Denmark Estonia Finland France Germany Hungary Iceland Japan New Zealand Norway Poland Portugal Slovak Republic Slovenia Spain Sweden Switzerland 7

8 Which Countries Are Appropriate Comparators? Total Health Expenditure per Capita, $U.S. (PPP), 29 Selected Countries, 2009 United States Norway Switzerland Netherlands Luxembourg CANADA Denmark Austria Germany France Belgium Sweden Iceland United Kingdom Australia(*) Finland Spain New Zealand Japan(*) Slovenia Portugal(*) Czech Republic Slovak Republic Korea Hungary Poland Estonia Mexico Turkey(*) Note * Data for Source OECD Health Data 2011, June edition. $5,352 $5,144 $4,914 $4,808 $4,363 $4,348 $4,289 $4,218 $3,978 $3,946 $3,722 $3,538 $3,487 $3,445 $3,226 $3,067 $2,983 $2,878 $2,579 $2,508 $2,108 $2,084 $1,879 Average: $3,343 $1,511 $1,394 $1,393 $918 $902 $7,960 $0 $1,000 $2,000 $3,000 $4,000 $5,000 $6,000 $7,000 $8,000 $9,000 8

9 Positive Correlation Between GDP Growth and Total Health Expenditure Growth, 29 Selected OECD Countries Average Growth, 2000 to 2009 Annual Average Real Growth In Total Health Expenditure Per Capita 12% SVK 10% KOR 8% POL EST 6% NZL NLD GBR CZE TUR 4% 2% DNK BEL ESP OECD FIN US CAN SWE MEX NOR AUS JPN LUX FRA DEU AUT CHE PRT ISL HUN SVN 0% 0% 1% 2% 3% 4% 5% 6% Annual Average Real Growth In GDP Per Capita OECD Health Data

10 Health Spending Per Capita ($US PPP) Health Spending Per Capita and Share of Population Aged 65 and Over, 29 Selected OECD Countries, USA MEX TUR NOR CHE LUX NLD CAN DNK AUT FRA BEL ISL AUS GBR SWE FIN NZL ESP SVN PRT SVK CZE KOR HUN POL EST DEU JPN Share of Population Aged 65 and Over Source :OECD Health Data 2011

11 SHA Classifications The SHA classification consists of three core dimensions: health care by functions of care (HC); providers of health care services (HP); and Financing schemes/agents(hf) Standard tables cross-classify expenditure between the dimensions. They include: Expenditure on health by function and provider (HCxHP) Expenditure on health by provider and financing schemes/agents (HPxHF) Expenditure on health by function and financing schemes/agents (HCxHF) 11

12 Some Countries Unable to Allocate to Core Variables Percentage of Core Expenditure Variables Reported AUS, AUT, CAN, CZE, DNK, EST, FIN, FRA, DEU, HUN, JPN, KOR, LUX, NZL, NOR, POL, PRT, SVK, SVN, ESP, SWE, USA 100% BEL, ISL, ISR, CHE 88% ITA, NLD 63% GRC, IRL, MEX, GBR 38% CHL, TUR 25% 0% 20% 40% 60% 80% 100% OECD Health Data

13 How Closely Do Countries Adhere to Definitions? Number of Countries Reporting Expenditures, by Core Variable Total Expenditure 34 Investment 29 Long-Term Care Curative & Rehab Care Inpatient Care Outpatient Care Pharmaceuticals Prevention/Public Health OECD Health Data

14 Historical Comparability: Health Expenditure Time-Series Australia Austria Belgium Canada Czech Republic Denmark Estonia Finland France Germany Hungary Iceland Japan Korea Luxembourg Mexico Netherlands New Zealand Norway Poland Portugal Slovak Republic Slovenia Spain Sweden Switzerland Turkey United Kingdom United States First Year of Data Reported According to SHA OECD Health Data

15 Future Items to Monitor Release of OECD Health Data 2012 (late June 2012) Estimating expenditure by disease, age and gender using a top down approach, under the SHA framework Development of output-based health specific purchasing power parities Improving the comparability and availability of private health expenditure Results of pilot projects on the feasibility of implementing specific parts of SHA 2.0 pertaining to LTC, OTC medicine and capital expenditures (Dec. 2012) 15

16 Thank you! Christopher Kuchciak

17 International Comparisons in Cancer Care Heather Bryant, MD, PhD VP Cancer Control Canadian Partnership Against Cancer CAHSPR May 2012

18 What to measure? Board Executive team Medical/Clinical Leadership Big dots Clinical and Corporate Programs Little Dots Heenan, M, Kahn, H & Binkley, D (2010) From Boardroom to Bedside: How to Define and Measure Hospital Quality Healthcare Quarterly, 13(1):

19 For National Cancer Control National Outcome Indicators Big Dots National/Province Driver Indicators Medical/Clinical/Program Indicators at Province Level Clinical and Corporate Quality and Process Indicators Little Dots

20 Common big dots in cancer comparisons Incidence and mortality built on long history of registries in developed countries, and an ongoing expansion of registries through work of International Agency for Research in Cancer (IARC; a WHO affiliate) Survival closer to connecting the dots between diagnosis and death Screening rates Because screening one of few interventions organized at population level, is one of few interventions with uniform and reliable data to compare All of above are in OECD reports and other international comparisons 20

21 Even with big dots, you can get a good picture of where things are

22 International comparisons in incidence/mortality CI5: Cancer Incidence in 5 continents available through IARC; Canada an early contributor 8 provincial registries represented Globocan: A more recent effort to impute likely incidence numbers in countries where cancer registration incomplete 22

23 Globocan liver cancer rates 23

24 Globocan grouped data 24

25 IARC CI5 trends, liver cancer CI5 data excellent and easy to manipulate to get international comparisons However, generally quite out of date due to data collection and cleaning lags Incidence may have more gaps than mortality 25

26 Cancer survival Is a function of stage at diagnosis and quality of treatment Usually disease-specific survival not calculable on populations ( sticky diagnosis and slippery linkage ) Relative survival compares survival of those diagnosed with cancer to mortality of general population: common international comparator 26

27 Survival comparisons EUROCARE: relative survival analysis in European countries: started with diagnosis years in 1970 s and is updating (most recent dx year 2002 at present) Resulted in awareness of relatively lower survival in UK compared to Western European countries CONCORD: 31 countries on 5 continents, published in 2008 (Coleman et al, Lancet Oncology) diagnosis ICBP: consortium of developed countries with national availability of health care with later modules to attempt to explain differences (Lancet, 2011) 27

28 CONCORD study: Colorectal (men) diagnosis 4 th for women and 6 th for men Survival in Canada is the same as the median survival for industrialized countries (56%) (is now higher about 63%) Presence or not of screening influences RS Surgical factors influence More recently, chemotherapy has a role in survival Source: Coleman et al. 2008

29 Rationale for ICBP The UK was aware that 5-year survival for cancer was lower in the UK than for many western European studies Early comparisons indicated that most of the difference may be in the first year of survival Desire was to do comparisons with like countries, using conditional survival to see whether this did, in fact, account for the most of the difference

30 International comparisons of cancer survival International Cancer Benchmarking Project Canada, Australia, Denmark, Norway, Sweden and UK colorectal, lung, breast, ovary 2.4 million individuals 1 and 5 year RS and 5 year Conditional survival Patients diagnosed 1995 to 2007 with follow-up to 2007 Alberta, BC, Manitoba and Ontario participated from Canada

31 ICBP: Colorectal Average age at diagnosis was age 71 Similar improvement in RSR six countries over time Most of the difference in RSR between countries occurred during the first year Larger inter-country differences in survival for patients aged over 65 Source: Coleman et al. 2010

32 Conditional survival, ICBP, colorectal cancer ICBP Country Australia 77.7 Canada 76.4 Norway 75.4 Sweden 74.8 Denmark 72.1 UK diagnosis years; colorectal cancer conditional survival 32

33 OECD Good organization for comparison base as most countries have similarly successful economies but different health systems Big dot indicators mortality and survival for major cancers, and screening uptake for breast, cervix, colorectal 33

34 OECD cervical screening rates Source of data varies among countries: Canada is self-report Recent analysis has shown self-report to be quite comparable to administrative data analyses (Doyle et al, 2011) 34

35 Shortcomings in international comparisons Relatively limited data on the patient care experience: infrastructural comparisons (PET scanners or radiotherapy unit per population base, for example) exist but don t reflect appropriate utilization CPAC just beginning to publish such data in Canada for interprovincial comparison; ability to do so a partnership with CPAC and provincial agencies/registries (available cancerview.ca Systems Performance Report) 35

36 Stage I or II Breast Cancer Patients Receiving RT Following Breast Conserving Surgery 36

37 Stage II or III Rectal Cancer Patients Receiving RT Preceding Resection 37

38 Patient perspective on indicators Heal me Don t hurt me Be nice to me Treat me quickly 38

39 Thank you!

40 Jean-Frédéric Levesque, CAHSPR 2012 International Comparisons: Adding Value to the Canadian Health Care System Primary Care Jean-Frédéric Levesque, MD, PhD Scientific Director, Health Systems Analysis and Evaluation, Institut National de Santé Publique du Québec CAHSPR Tuesday, May Montréal Concurrent Sessions Stream A - 10:30am 11:45am

41 objective «to showcase some international comparisons available to compare primary healthcare and discuss some challenges and opportunities.» 41 Jean-Frédéric Levesque, CAHSPR 2012

42 Primary healthcare? the care provided by certain clinicians general practitioners, nurses, health professionals a set of activities whose functions define the boundaries of primary care such as curing or alleviating common illnesses and disabilities a level of care or setting an entry point to a system a set of attributes care that is accessible, comprehensive, coordinated, continuous, and accountable) or care that is characterized by first contact, accessibility, longitudinality, and comprehensiveness (Starfield (1992) 42 Defining Primary Care: An Interim Report (IOM, 1994). Jean-Frédéric Levesque, CAHSPR 2012

43 Comparing PHC results/outcomes? OECD indicators on primary healthcare Lifestyle, prevention and screening Avoidable hospitalisation/mortality ambulatory care sensitive conditions An archetype perspective to group systems funding, governance and coverage Utilisation and payment measures Limited set of indicators, comparability issues CIHI PHC indicators as a framework for developing a broader set of comparable indicators at the international level? 43 Jean-Frédéric Levesque, CAHSPR 2012

44 Average Annual Number of Physician Visits per Capita, JPN* GER FR AUS OECD Median NETH CAN* UK DEN NZ** SWIZ** US* SWE * ** Source: OECD Health Data 2011 (June 2011).

45 Percent Cervical Cancer Screening Rates, 2006 Women Ages US UK SWE NOR CAN* FR* NZ NETH AUS * Based on survey data; all other countries based on program data. 45 Source: Health Care Quality Indicators Project, OECD Jean-Frédéric Levesque, CAHSPR 2012

46 Jean-Frédéric Levesque, CAHSPR 2012 Integration Hospitalisation Costs ER use Measuring health systems impacts Measuring social and population impacts Population coverage Mortality Adverse events Satisfaction and confidence

47 Comparing PHC processes and outputs? Commonwealth fund s International health policy surveys Population experience of care Sicker adults experience of care Reporting and rating Challenges related to taking into account the cultural and social characteristics as well as expectations towards health systems Increasing participation at the sub-country level? 47 Jean-Frédéric Levesque, CAHSPR 2012

48 Emergency room visit was for a condition that could have been treated by the regular doctor, countries and provinces, 2010 Canada Ontario Quebec 46 % 49 % 46 % 54 % 51 % 54 % France Germany Norway Australia New Zealand Netherlands Sweden United Kingdom Switzerland United States 23 % 26 % 32 % 34 % 35 % 36 % 39 % 41 % 42 % 49 % 77 % 74 % 68 % 66 % 65 % 64 % 61 % 59 % 58 % 51 % 0 % 10 % 20 % 30 % 40 % 50 % 60 % 70 % 80 % 90 % 100 % Yes No Among respondents that had a hospital emergency room visit in the past 2 years Jean-Frédéric Levesque, CAHSPR 2012

49 Feet examined in the past year (respondents with diabetes), countries and provinces, 2008 Jean-Frédéric Levesque, CAHSPR 2012

50 Measuring organisation al outputs Measuring patients outcomes Measuring health systems impacts Reception of services Adverse events Access and coverage Continuity and patient-centred Measuring social and population impacts Perceived health Functioning Health-related quality of life Morbidity and mortality

51 Comparing PHC structure and organisation? Commonwealth fund s International health policy surveys Survey of primary care providers The challenging link with models and reforms Disentangling the provider and organisational levels Developing organisational typologies and tailoring some questions to contexts Potential developments in organisational surveys Évolution, BEACH, TranforMed, QUALICOPC Jean-Frédéric Levesque, CAHSPR 2012

52 52 Origin Name of the Survey/ Project Acronym # Form # Items Respondent Commonwealth Fund International Survey of Primary Care Doctors (2009) ISPCD 1 40 PR Europe Physician Questionnaire (Evaluating costs and quality of primary care in Europe) QUALICOPC 1 62 PR Australia Bettering the Evaluation And Care of Health BEACH 1 20 PR New-Zealand National Primary Medical Care Survey NatMedCa 2 42 PHCO & PR United Kingdom Improving the delivery of care for patients with type 2 diabetes IDCP2D 2 35 PHC staff & PHCO United States Canada Ontario (Canada) Québec (Canada) Nova-Scotia (Canada) National Ambulatory Medical Care Survey NAMCS 3 77 PHCO Primary Care Practice Site Survey PCPSS 1 35 PHCO Physician Practice and Quality of Care Survey PPQCS 1 35 PHCO Survey of Organizational Attributes for Primary Care SOAPC 1 21 PHC staff Methods for evaluating practice change toward a patient-centered medical home TransforMED PHCO National Study of Physician Organizations and the Management of Chronic Illness II NSPOII PHCO National Family Physician Workforce Survey NFPWS PR National Physician Survey NPS PR Comparison of Models of Primary Health Care in Ontario COMP-PC 1 20 PHCO Improving Measurement for Evaluation in Primary Health Care IMEPHC 1 20 PHCO Continuity of Primary Care in Quebec CPCQ 3 85 PHCO & PR Assessing the evolution of primary healthcare organizations and their performance Evolution 1 65 PHCO Clinicians perception of organizational readiness for change CPORC 1 36 PHC staff Primary Care Organization Survey PCOS-NS 2 97 PHCO Jean-Frédéric Levesque, CAHSPR 2012

53 Jean-Frédéric Levesque, CAHSPR 2012 Hours worked by week in the medical practice, countries and provinces, 2009 Canada Ontario Québec 35.1 Germany United States France Netherlands United Kingdom New Zealand Australia Norway Sweden Italy Hours worked by week in the medical practice

54 Jean-Frédéric Levesque, CAHSPR 2012 Generate a list of all medications taken by an individual patient, countries and provinces, 2009 Canada 35 % 17 % 48 % Ontario 41 % 15 % 44 % Quebec 22 % 15 % 63 % United Kingdom 90 % 5 % 5 % Australia 71 % 21 % 8 % Netherlands 70 % 17 % 13 % Norway 60 % 14 % 26 % Germany 59 % 14 % 27 % New Zealand 57 % 28 % 15 % Italy 53 % 24 % 22 % United States 49 % 19 % 32 % Sweden 45 % 13 % 42 % France 45 % 19 % 36 % 0 % 10 % 20 % 30 % 40 % 50 % 60 % 70 % 80 % 90 % 100 % Easy Somewhat Difficult Difficult / Cannot Generate

55 Canada s respondents Network of practices FTE doctors Practice setting Electronic records Other providers Jean-Frédéric Levesque, CAHSPR 2012 Mean number of non-physician FTE health care providers, Canada s respondents (practice), 2009 Solo practice 0.4 Group practice 2.3 Family medicine group 4.7 Local com. health centre 15.6 Hospital / Other 10.3 Less than to less than and more 11.8 Yes 7.6 No 3.9 Yes 6.1 No 2.8 Yes No Mean number of non-physician FTE health care providers

56 Jean-Frédéric Levesque, CAHSPR 2012 Measuring organisational outputs Measuring provider outcomes Measuring patients outcomes Measuring health systems impacts Measuring organisational functioning Measuring social and population impacts Work satisfaction Provider health Perception of problems Career pathways Team climate Collaboration Coordination Clinical governance

57 Jean-Frédéric Levesque, CAHSPR 2012 Measuring organisational outputs Measuring provider outcomes Measuring patients outcomes Measuring health systems impacts Measuring organisational functioning Measuring social and population impacts Organisational culture Resources Structure and governance Clinical coverage Measuring organisations

58 Jean-Frédéric Levesque, CAHSPR 2012 Measuring organisational outputs Measuring provider outcomes Measuring patients outcomes Measuring health systems impacts Measuring organisational functioning Measuring social and population impacts Measuring physician panel composition Measuring patients panel composition Vision and values Skills and competencies Expectations and goals Measuring organisations Vision and values Skills and competencies and capacity Expectations and goals

59 Integrated measurement systems? CIHI development of instruments project Patients experience, provider survey, practice survey and administrative databases indicators Nivel - QUALICOPC 31 countries patients, providers and organisational surveys Comparability vs accuracy... Challenges of going beyond the single study timeframe and transform them into routine systems 59 Jean-Frédéric Levesque, CAHSPR 2012

60 Health Administrative Databases Patients surveys Measuring organisational outputs Measuring provider outcomes Measuring patients outcomes Measuring health systems impacts Measuring organisational functioning Measuring social and population impacts Measuring physician panel composition Measuring patients panel composition Measuring organisations Provider surveys Organisational surveys Jean-Frédéric Levesque, CAHSPR 2012

61 Thank you!

62 G. Ross Baker, Ph.D. University of Toronto CAHSPR Conference 29 May 2012

63 Patient safety is the reduction and mitigation of unsafe acts within the health-care system, [and] the use of best practices shown to lead to optimal patient outcomes Canadian Patient Safety Dictionary, 2003 Adverse events can be defined as physical or psychological harm to patients arising from health system contact; such harm is not associated with the disease process, comorbid conditions, or expected treatment outcomes Miller, et al., 2001

64 Health System Structures and Policy Management Processes (latent errors) Interventions Intervening Variables (e.g., Culture) Clinical Processes (active errors) Patient Outcomes Source: Modified from Brown, Hofer et al., 2008; and based on Donabedian, 1966

65 Measure Focus International Scope Measures of Adverse Events, including Global Trigger Tool Adverse events, in acute care and other sectors Studies done in many countries Patient Safety Indicators (PSIs) Adverse events in acute care Developed in US; International comparisons in OECD countries underway HSMR Acute Care Developed in England, now used in Canada, US, Sweden, Netherlands and elsewhere

66 Measures of the scale and impact of patient safety events pose important challenges: Different measures of the same construct vary substantially in prevalence Reporting systems are a valuable resource but often difficult to implement and maintain Reporting cultures vary widely between organizations and across health care systems Routine data offer an opportunity for comparisons but these systems may vary in their sources and extent to which data are valid and reliable

67 Country Charts Reviewed Year Incidence of AE Preventable? Canada % 37% Denmark 1, % 40.4% New Zealand 6, % 37% England 1, % 50% Australia 14, % 51% USA (Utah & Colorado) 15, % - USA (NY) 30, % 58%

68 Tsang and colleagues (2012) identified 124 citations related to patient safety measures generated from routinely collected data 80% published since 2005 Although they covered a wide range of events, many were complications associated with surgery or obstetrical care

69 The OECD convened an international expert panel which examined 59 indicators and endorsed 21 for international use on following criteria: Impact on health Policy importance Susceptibility to influence by health system change Face validity Content validity Data availability Reporting burden Twelve of the 21 indicators were derived from the AHRQ Patient Safety Indicators set

70 Anaesthesia complications Death in low mortality DRGs Decubitus ulcer Failure to rescue Foreign bodies left during surgery Iatrogenic pneumothorax Infection due to medical care Post op hip fractures Post op hemorrhage or hematoma Post op physiologic or metabolic derangement Post op respiratory failure Post op thromboembolism Post op septicemia Post op abdominopelvic wound dehisence Accidental puncture or laceration Transfusion reaction Birth trauma Obstetric trauma

71

72 Patient safety indicators have been validated in both ICD-10 and ICD-9 coded data but variations in rates suggest that these measures may be influenced by coding practices and data storage issues Recent study of a subset of 7 PSIs indicates that non-obstetrical indicator rates are positively correlated with the mean number of secondary diagnoses This finding suggests that complication related diagnoses are more likely to be reported in areas where routine coding practices require more thorough review of clinical documentation and more complete coding Or more simply where clinical documentation is more complete Variations in coding practices and responsibility across countries require cautious use of PSIs and exploration of the source of variation

73 Differences in health system may confound comparisons of patient safety indicators Average LOS in OECD countries varies from 3.5 to 7.8 days patients with longer stays have greater risk exposure to patient safety events

74 English researchers have examined the application of 9 PSIs to English hospital data Positive cases were associated with higher mortality, greater risk of unplanned readmission and longer length of stay Dr Foster Intelligence is working with a group of NHS hospitals to improve the accuracy of admissions data and to assess their usefulness in measuring and monitoring patient safety

75 Hayes, et al study of the implementation of the surgical safety checklist in 8 hospitals in 8 cities demonstrated the impact of a similar intervention in differing contexts Overall the death rate was reduced from 1.5% to 0.8 % (p=0.003) and complications from 11.0% to 7.0% after introduction of the checklist

76 Haynes, et al., NEJM 2009

77 Relatively few measures have been developed to provide international comparisons of patient safety The AHRQ PSI indicators have been a major focus of interest However, these are largely surgical and obstetrical and are all acute care Additional measures in other sectors and cross-sector are needed to provide international comparisons Validated patient safety indicators that can be routinely collected and analyzed are an important contribution to patient safety efforts Measures that can be applied cross-nationally offer opportunities to focus on system practices that may support higher quality care and outcomes

78 Questions for our panelists? 78