Appendix 1: About the geographical information system analysis

We used a geographical information system (GIS) to calculate the proportion of a district health board's population for which the district health board (DHB) had planned availability to an after-hours service within 60 minutes' drive.

We acknowledge the assistance of Critchlow Limited in carrying out the GIS analysis on our behalf.

A simplified summary of the analysis carried out for each DHB is:

1. identify the after-hours services that are available and when they are available – the source of this information was each DHB's after-hours plan and rural roster information requested through DHBs;
2. record the availability of after-hours services into time slots and days (21 different time and day slots) – the time slots were selected based on service use and common open and closing times;
3. obtain the physical addresses of the after-hours service providers – achieved using various online information sources and by telephone calls;
4. verify the information obtained in steps 1-3 with the DHB (all DHBs confirmed that the information was complete and accurate);
5. using a GIS, code each physical address to a corresponding geographical reference – such as latitude and longitude co-ordinates;
6. assume that people use the fastest route on the roading network to travel from their home to the nearest after-hours service regardless of whether the service is located inside or outside of a DHB's boundary and complete steps 7-10;
7. using a GIS, including the New Zealand roading network and information on actual average road travel speeds (based on 1,831,034 readings from vehicles), calculate the geographical area that has access to each after-hours provider within 60 minutes' drive – this was done for each time slot and day;
8. using a GIS, identify and map meshblocks (see Figure 8) into three categories using the information from step 7 – within 60 minutes' drive for all time slots and days, for some time slots and days, or for no time slots and days;
9. using a GIS (including meshblock, postal address, and census information), calculate the population of the meshblocks always within 60 minutes' drive (that is, for all time slots and days) – in urban areas the population of each meshblock is included when the centre of the meshblock falls within 60 minutes' drive, and in rural areas the population of a meshblock is included in proportion to the number of address points in the meshblock falling within 60 minutes' drive;
10. using 2006 Census information and the results from step 9, calculate the percentage of each DHB's population that is able to access after-hours services within 60 minutes' drive; and
11. run through steps 7-10 again, assuming that people use the fastest route in the roading network to travel from their home to the nearest after-hours service located inside a DHB's boundary.

A more detailed technical document outlining the information sources, assumptions, and analytical methodology is available on request. Readers interested in obtaining a copy can email their request to [email protected].