Finland, Iceland, the United Kingdom (England) and Singapore are regularly linking most of their national health care datasets for statistics and research

Finland, Iceland, the United Kingdom (England) and Singapore reported that 90% or more of their key national health care datasets were being linked to other health and health care datasets to regularly monitor health care quality or system performance or to produce other approved statistics or research studies (Table 2.1).

Monitoring pathways of care and the outcomes that result is essential to understanding and improving health care quality and performance. Seldom is all of the data needed to understand a pathway available within a single dataset and data linkages are needed to bring together data across the continuum of care for the same patients. Dataset linkages also enable key characteristics of individuals, such as socio-economic status, health behaviours and environmental exposures to be linked to health care datasets to provide a new understanding about the impact of health risks and inequalities on individuals’ health and health care.

Most countries reported developing indicators for health care quality or health system performance monitoring from hospital in-patient data (19 countries) and cancer registry data (18 countries); followed by mental hospital in-patient data (15 countries) and population health survey data (14 countries). About half of countries have gone further and are developing indicators across the continuum of health care.

Indeed, all of the countries with personal hospital in-patient data reported that the data are used for approved data linkage projects (Table 2.6). Similarly, the vast majority of countries reported data linkage projects involving their datasets across the spectrum of national health data, with the exception of survey data. Population health survey and patient experiences survey data are less likely to be reported as having identifying variables and only half of countries with these data reported having linked the data for approved projects.

Countries were asked if their data are linked on a regular-basis such that a project involving the linkage of the data is usually underway. Cancer registries were the most frequently reported as involved in a regular data linkage (17 countries), followed by hospital in-patient data (16) and mortality data (15).

Hospital in-patient data are a key input to most programmes of quality and performance monitoring. Among the 19 countries where hospital datasets contain a unique patient identifying number, 17 could link these data to mental hospital in-patient data, 15 to mortality data, 14 to population census or registry data, 13 to emergency health care data, 13 to cancer registry data, 11 to prescription medicines data and 10 to primary health care and long-term care data. This technical capacity to link is because the datasets share the same unique patient identifying number. Fewer countries link these data on a regular basis for health care quality monitoring. Very few countries include linkages of their mental hospital in-patient data, emergency health care data, primary care data, prescription medicines data, and long-term health care data in their programmes of health care quality and system performance monitoring.

Table 2.6. Just over half of countries with national datasets are regularly linking the data to monitor quality or

health system performance

Hospital

in-patient

data

Mental

hospital

in-patient

data

Emergency health care data

Primary care data

Prescription

medicines

data

Cancer

registry

data

Diabetes

registry

data

Cardiovascular disease registry data

Mortality

data

Formal long-term care data

Patient reported health care outcomes data

Patient experiences survey data

Population health survey data

Population census or registry data

Dataset contains records for patients (persons)1

20

18

14

10

14

19

7

8

21

11

7

8

19

19

Data set contains a number that uniquely identifies the patient (or the person) that could be used for an approved data linkage

19

18

13

10

12

16

6

8

17

10

4

5

12

14

Data set contains the same identifying number as is used for hospital in-patient data2

na

17

13

10

11

13

6

8

15

8

4

6

8

11

Dataset contains identify ng variables (such as name, sex, birth date, address) that could be used to link these data to another set of data

16

15

11

9

11

15

5

6

17

10

4

5

11

13

Record linkage projects are conducted with these data

20

16

15

10

13

18

5

7

19

10

5

4

9

14

Record linkage projects are conducted with these data on a regular basis

16

13

10

4

10

17

2

5

15

7

4

2

7

9

Record linkage projects are conducted with these data on a regular basis for health care quality or system performance monitoring

13

7

6

2

3

12

1

4

11

2

1

0

2

2

  • 1. Each row may be one treatment/visit/admission that can be grouped by person.
  • 2. Where pseudonymisation algorithms are used for linkage, they would need to be consistent across datasets.

Source: Authors own calculations based on the results of this study.

Thirteen countries are regularly linking data from four national datasets: hospital and mental hospital in-patient data, cancer registry data and mortality data (Table 2.7). Ten of them are also regularly linking emergency care data, eight prescription medicines data, six long-term care data and three primary care data.

Table 2.7. Thirteen countries are linking data across the pathway of care

A

B

C

D

E

Regularly linking hospital inpatient, mental-hospital inpatient, cancer registry data and mortality data

Linking datasets in A + emergency care data

Linking datasets in A + prescription medicines data

Linking datasets in A + long-term care data

Linking datasets in A + primary care data

Canada

Canada

Canada

Canada

Korea

Czech Republic

Israel

Denmark

Finland

Singapore

Denmark

Korea

Finland

Israel

United Kingdom (Wales)

Finland

New Zealand

Korea

Korea

Israel

Norway

New Zealand

Singapore

Korea

Singapore

Sweden

United Kingdom (Wales)

New Zealand

Sweden

United Kingdom (Scot. & Wales)

Norway

United Kingdom (Eng., Scot. & Wales)

Singapore

Sweden

UK (Eng., Scot. & Wales)

Source: Authors own calculations based on the results of this study.

Countries provided examples of the purpose of the regular data linkages they are undertaking. Key reasons include to develop health care quality and system performance indicators including OECD quality indicators; to measure the co-ordination of care and health care pathways and outcomes; for estimates of compliance to national care quality guidelines; for indicators of health care utilisation and its cost; for measures of disease prevalence; and to measure health and health care use by socio-economic status. Linkage is also routinely necessary to assure the quality, completeness and validity of national datasets and to conduct medical and health services research projects.

These 13 countries also stand out for the number of national datasets regularly included in data linkage projects for health or health care monitoring, statistics or scientific research (Table 2.8). Iceland is also conducting data linkage projects with the majority of its key national datasets. Countries with few key national datasets regularly involved in a data linkage projects are Ireland, Italy and Switzerland. In Japan and Turkey there are no regular dataset linkages.

Table 2.8. Seven countries are linking seven or more key datasets on a regular basis for statistics or research

7+ key national datasets

Canada, Finland, Iceland, Israel, Korea, Norway, Singapore, Sweden, United Kingdom (England, Scotland and Wales)

5-6 national datasets

Denmark, New Zealand, Netherlands, United States

3-4 national datasets

Czech Republic, Spain

1-2 national datasets

Ireland, Italy, Switzerland

0 national datasets

Japan, Turkey

Source: Authors own calculations based on the results of this study.

 
Source
< Prev   CONTENTS   Source   Next >