A 21st Century Data Strategy for Health and Health Care Surveys: Their Unique Contribution to Improving Health Outcomes and Reducing Disparities Arlene.

A 21st Century Data Strategy for Health and Health Care Surveys:

Their Unique Contribution to Improving Health Outcomes and Reducing Disparities

Arlene S. Bierman MD MSOWHC Chair in Women’s Health

St. Michael’s Hospital, University of Toronto

AHRQ Annual MeetingSeptember 29 , 2010

2

Health and Health Care SurveysEssential Data for Improving Health Outcomes

Assessing, Improving, and Monitoring• Health System Performance• Population Health• Health Disparities

Identifying• Individuals, Populations, and Communities at Risk

Benchmarking Conducting International Comparisons

3

Health and Health Care SurveysUnique Contributions

Patient Reported Outcomes• Health and Functional Status• Physical and Mental Health

Health Behaviors and Risk Factors Patient Experiences with Care Non-Medical Determinants of Health Health Needs of Diverse Populations

4

Actionable Data for Improvement

The POWER Study (Project for an Ontario Women’s Health Evidence-Based Report) is providing actionable data to help policymakers and providers to improve the health of and reduce inequities among the women of Ontario.

http://www.powerstudy.ca

5

Community-Engaged Research POWER Study Roundtables

– Inform indicator selection and interpretation– Increase uptake of findings

Consumers: representatives of community based organizations and associations

Providers: Clinicians, Hospitals, Community Health Centres

Policymakers: Government, Regional Health Authorities, Public Health, Health Data Agencies

6

Overall Population

Women Men

Income

Education Ethnicity

Geography Geography

EthnicityEducation

Income

Assessing Equity

7

Patient Reported Outcomes

8

Age-specific percentage of adults aged > 25 years who reported activities were prevented due to pain or discomfort, by sex and

annual household income, Ontario, 2000/01

26

35

2527

16

27

16

26

15

23

1113

107

11*

18*

0

10

20

30

40

Women 25–64 Women 65+ Men 25–64 Men 65+

Sex and age group (years)

Per

cent

age

(%)

Low income Low er middle income Middle income Higher income

Data source: Canadian Community Health Survey cycle, 1.1

Age-standardized percentage of adults age ≥ 25 with CVD who reported that their current health was somewhat or much worse than

their health one year prior, by sex and annual household income, 2005

9

3531

2728

221920*

33*

0

10

20

30

40

50

Low Lower middle Middle Higher

Annual household income

Per

cent

age

(%)

Women Men

Data source: CCHS, Cycle 3.1 * Interpret with caution due to high sampling variability (coefficient of variation

16.6–33.3)

10

Risk Factors

11

Age-standardized percentage of women aged > 25 who reported health behaviors that increase the risk of chronic

diseases, by education level, Ontario, 2005

6560

55

28

57 5646

26

51 49 46

21

4842

34

8

0

20

40

60

80

100

Physical inactivity* Inadequate fruit andvegetable intake**

Overweight or obese*** Smoking

Education level

Less than secondary school graduation Secondary school graduation

At least some post-secondary school Bachelor's degree or higher

Data source: Canadian Community Health Survey cycle, 3.1 *Physical activity index was less than 1.5 kcal/kg/day** Less than five servings per day***Body Mass Index (BMI) >greater than or equal to 25 (calculated from self-reported height and weight) ^Daily or occasional smokers

12

Age-standardized percentage of adults aged 25 years and older who reported being current smokers,

by sex and ethnicity, Ontario, 2005

39

13

22

43

18 18

25 25

4*5*10*

19*

0

10

20

30

40

50

Aboriginal** Black South and WestAsian, Arab

East andSoutheast Asian

Other*** White

Ethnicity

Per

cent

age

(%)

Women Men

Data source: Canadian Community Health Survey 3.1 *Interpret with caution due to high sampling variability (coefficient of variation 16.6–33.3)**Only includes off-reserve Aboriginals (North American Indian, Metis, Inuit) ***Includes Latin American, other racial and multiple racial origins.

Age-standardized percentage of adults aged 25 and older

who reported being a daily or occasional

smoker, by sex, education level and

Local Health Integration Network,

Ontario, 2005

13

14

Age-standardized percentage of adults age ≥ 25 with CVD who reported health behaviors that increase risk for chronic diseases,

by sex and risk behaviour, Ontario, 2005

Data source: CCHS, Cycle 3.1 ^ Physical Activity Index of < 1.5 kcal/kg/day** Daily consumption of less than five servings of fruits and vegetables¥ Body Mass Index (BMI) ≥25, calculated from self-reported height and weight$ Current smokers (daily or occasional)

67

4854

14

5158

66

16

0

20

40

60

80

100

Physical inactivity Inadequate fruit andvegetable intake**

Overweight or obese¥ Smoking$

Health behaviour

Per

cent

age

(%)

Women Men

15

Social Determinants of Health

16

25

14

6

24

11

41* 1*

0

10

20

30

Low Lower middle Middle Higher

Annual household income

Per

cent

age

(%)

Women Men Data source: Canadian Community Health Survey 3.1 ^ Refers to people who reported that they did not have enough to eat, worried about there Not being enough to eat or did not eat the quality or variety of foods desired due to a lack of money*Interpret with caution due to high sampling variability (coefficient of variation 16.6–33.3)

Age-standardized percentage of adults aged 25 and older who reported food insecurity^, by sex and annual household income, Ontario, 2005

17

AccessPatient Experiences with Care

Percentage of adults aged 25 and older who reported no difficulties making an appointment for an urgent, non-emergent health problem,

by sex and neighbourhood income quintile, Ontario, 2006–08^

18

76 81 8579 84

77 79 84 86 88

0

20

40

60

80

100

Q1 (Lowest) Q2 Q3 Q4 Q5 (Highest)

Neighbourhood income quintile

Per

cent

age

(%)

Women Men

Data sources: Primary Care Access Survey (PCAS), Waves 4–11; Statistics Canada 2006 Census^ October 2006–September 2008

Percentage of adults aged > 25 who reported being very satisfied with their experience of getting an appointment for a regular check-up, by

sex and ethnicity, 2006–08^

19

73

5247 49

58 6170

50

40

6965

X0

20

40

60

80

100

Aboriginal** Black South and WestAsian, Arab

East andSoutheast Asian

Other*** White

Ethnicity

Per

cent

age

(%)

Women Men

Data source: Primary Care Access Survey (PCAS), Waves 4–11^ The survey period was from October 2006–September 2008X Suppressed due to small sample size ** Includes North American Indian, Metis, Inuit*** Includes El Salvador, other European, other Central American, other South American, religion as ethnicity

Percentage of adults aged 25 and older who reported being very satisfied with their experience getting an appt for a regular check-up,

by sex and length of time since immigration, 2006–08^

20

41

61 60

42

61 65

0

20

40

60

80

100

0 - 9 10+ Canadian born

Time since immigration (years)

Per

cent

age

(%)

Women Men

Data source: Primary Care Access Survey (PCAS), Waves 4–11^ October 2006–September 2008

Percentage of adults aged 25 and older who reported being very satisfied with their experience of getting appt for a regular check-up,

by sex and language spoken most often at home, 2006–08^

21

6068

49

6571

51

0

20

40

60

80

100

English only, English with others French only Neither English nor French (other)

Language spoken most often at home

Perc

enta

ge (%

)

Women Men

Data source: Primary Care Access Survey (PCAS), Waves 4–11^ October 2006–September 2008

Percentage of adults aged 25 and older who did not visit a dentist in the past 12 months,

by sex and annual household income, Ontario, 2005

22

56

48

29

16

57 55

40

23

0

25

50

75

Low Lower middle Middle Higher

Annual household income

Pe

rce

nta

ge

(%

)

Women Men

Data source: Canadian Community Health Survey (CCHS), 2005 (Cycle 3.1 )

23

Quality of Care

Percentage of Ontarians aged 15 and older with probable depression who had a physician visit for depression,

by sex and age group

24Data sources: CCHS, Cycle 1.1; OHIP* Interpret with caution due to high sampling variability

27

4350

4638

47

26* 28*

0

20

40

60

80

100

15-24 25-44 45-64 65+

Age group (years)

Per

cent

age

(%)

Women Men

25

Age-standardized percentage of screen-eligible^ women who had at least one Pap test in the last three years,

by neighbourhood income quintile, 2004/05

6166 69 72 75

0

20

40

60

80

100

Q1(lowest) Q2 Q3 Q4 Q5(highest)Neighbourhood income quintile

Per

cent

age

(%)

Data sources: CytoBase; OCR; OHIP; RPDB; Canadian Institute for Health Information Discharge Abstracts Database (CIHI-DAD); Statistics Canada 2001 Census

^Women aged 18-70 with no history of cervical cancer or prior hysterectomy

26

Age-standardized percentage of women who had a Pap test that showed a low grade lesion^ who had a repeat Pap test or colposcopy

within 6 months of the initial abnormal test, by neighbourhood income quintile, 2004/05

42 43 44 45 47

0

20

40

60

80

100

Q1(lowest) Q2 Q3 Q4 Q5(highest)

Neighbourhood income quintiles

Per

cent

age

(%)

Data sources: CytoBase; OCR; OHIP; RPDB; CIHI-DAD; Statistics Canada 2001Census

^Atypical squamous cells of undetermined significance (ASCUS) or low-grade squamous intraepithelial lesion (LGSIL)

27

Quality of Care:Medicare Health Outcomes Survey

Plan-level HEDIS diabetes indicators linked to patient-level HOS data.

Hierarchical linear models estimated the relationship between plan HEDIS performance on diabetes QIs and 2-year change in HOS physical and mental health scores.

Each 10% point improvement in plan performance on intermediate outcomes (ie, the proportion of well-controlled diabetes) was related to significant increase in the probability of being healthy for physical health scores (7 percentage point increase, P 0.05) and mental health scores (11 percentage point increase, P 0.01).

Source: Harman et al. Medical Care 2010

28

Identifying Populations at RiskCART Analysis

29

ACSC RATE: 4.9%

Sample CART Tree Profile

Total Population

N = 2,818,597

AGE > 49AGE <= 49

No ChronicDisabilities

Chronic Disabilities Present

Good to Excellent Mental Health

Fair to Poor Mental Health

Middle to High Income

Low Income

Employed

Unemployed

HEA

LTH

/CLI

NIC

AL

DET

ER

MIN

AN

TS

S

OC

IO-E

CO

NO

MIC

DETER

MIN

AN

TS

CART Risk Profiles

30

Sample Groups Formed by CART Analysis

Health +Socio-EconomicDeterminants

(Health Very Good to ExcellentAge < 50)Middle to High IncomeEmployed

(Health Very Good to Excellent Age < 50)Lower to Middle to IncomeNot Employed

(Health Good to ExcellentAge >= 50)Low Income

(Health Poor to AverageAge >= 50)Working Part Time Household Size 2 or lessLanguage-Non English

(Health Poor to AverageAge >= 50)Working Part Time Household Size 3 or moreLanguage-Non English

31

Data Linkages

Physician Claims Pathology Data Hospital Discharge Data Performance Data Other

– Census– Other Surveys– Lab Data– EMR?– All Payer Databases?

32

Future Directions: A 21st Century Data Strategy

Survey Development: Asking What Matters Fostering Data Linkages Oversampling of Diverse Populations Knowledge Translation (Translating Research

into Practice) Support Priority Setting, Inform Policy and

Practice, Monitor Progress Innovative Analyses and Pragmatic Trials Community Engagement

For more information, please contact us:

33

The POWER Study is funded by Echo: Improving Women's Health in Ontario, an agency of the Ministry of Health and Long-Term Care. This presentation does not

necessarily reflect the views of Echo or the Ministry.