Working with big health data. The Ministry of Health s role as an enabler and facilitator of safe access to data

Transcription

1 Working with big health data The Ministry of Health s role as an enabler and facilitator of safe access to data

2 Content The role of Analytical Services Ministry of Health in-house analyses New developments data sharing and the Integrated Data Infrastructure

3 Analytical Services What we do 1. Supply data to researchers and operational users across government, academia, the health sector and beyond 2. Advise on the use and interpretation of supplied data Expertise Metadata Conduit for data quality advice and investigation Conduit for coding advice and explanations 3. Protect patient privacy and confidentiality and Ministry of Health information security through safe management of access to data at patient and health event level

4

5 Ministry of Health in-house analyses Health Tracker A system that integrates data at patient level from multiple Ministry of Healthheld sources two components: A population spine Tracker indicators (conditions, costs, counts of service use) Virtual Diabetes Register Prevalence of diabetes in New Zealand constructed from indications of diabetes for specific patients in Ministry of Health datasets Hospital, outpatient, pharmaceutical, laboratory and mortality datasets are key sources Suicides amongst mental health service users

6 Prevalence rate 18% New Zealand Diabetes Prevalence Rates 1 January 2010: PHO enrolled population 16% 14% 12% New Zealand 10% 8% 6% 4% 2% 0% Age

7 50.00% VDR as at 31 Dec 2012 European/Other Indian Mäori Pacific people 45.00% 40.00% 35.00% 30.00% 25.00% 20.00% 15.00% 10.00% 5.00% 0.00%

8 Number and age-standardised rate of suicides, by service use, ages 10 to 64, 1 January to 31 December 2010 a Number Age-standardised rate b Deaths due to intentional self-harm Service users c Non-service users Total Deaths of undetermined intent Service users Non-service users Total Total deaths Service users Non-service users Total Notes: The definition of service user has changed this year to include only face-to-face contacts. (In previous years telephone contacts were included in this definition.) Equivalent figures for 2009: intentional self-harm: service users 155 (ASR 189.6), non-service users 304 (ASR 7.9), total 459 (ASR 12.3); undetermined intent: service users 6 (ASR 7.6), non-service users 10 (ASR 0.2), total 16 (ASR 0.4). Total: service users 161 (ASR 197.2), non-service users 314 (ASR 8.1), total 475 (ASR 11.9). aservice user denominator excludes service users with unknown age. Extracted on 4 July bage-standardised rate is per 100,000, standardised to the WHO population aged 0 64 years. cplease note that the 2010 ASR of service user suicides is higher (159= ASR) than that reported in the 2011 annual report (166= ASR) (This 2009 figure has been subsequently revised, please see above footnote.) This change in ratio is due to the change in methodology between the 2011 and 2012 publications as described above.

9 Statistics NZ s Integrated Data Infrastructure Source: Statistics New Zealand

10 Statistics NZ s Integrated Data Infrastructure (continued) Source: Statistics New Zealand

11 Adding health administrative data Two fold approach: 1. Pilot research project the impact of chronic conditions and acute health events on future earnings, benefit receipt and heath outcomes in the working age population Treasury led analysis Use of standard health ethics approval process National Collections datasets e.g. NMDS, cancer, mortality, outpatient events, pharmaceuticals, etc. 2. Decisions about ongoing sharing What datasets to supply? What governance model to operate? Where does ethical review of proposed uses fit?

12 Datasets supplied for the pilot Restrictions Age-limited cohort focussed on the working age population Aged at some point between 2002 and 2009 Data Flags indicating chronic conditions and acute events e.g. cancer, diabetes, coronary heart disease, gout, COPD, TBI, stroke and acute myocardial infarction Cancer registrations, mortality events, publicly funded hospital discharges, outpatient and ED events, laboratory claims, pharmaceutical dispensings, GMS claims, PHO enrolments For integration into this IDI in September, accessible from October 2014.

13 Next steps Supply Supply the same datasets for all NZ residents? Ongoing supply of data on a regular schedule? Expand the number of datasets supplied? Use Ministry of Health access to the IDI Encourage health and social sector researchers to use the information available to conduct cross-sectoral investigations