The proposed Geographic Classification for Health (GCH) will be based on the Statistics NZ Urban Accessibility Classification which is in turn based on the Statistical Standard for Geographic Areas Urban Rural Geography (SSGA18). The proposed GCH taxonomy is comprised of five categories, two urban and three rural, that reflect degrees of reducing urban influence and increasing rurality with respect to health and health services. The GCH will use these categories to classify all of NZ’s SA1s (small statistical areas which are the output geography for health data). Like the UAC, the proposed GCH is based on population size and population density, with drive time (to the edge of the urban area) used as the measure for reduced urban access / increasing rurality.
Three potential GCH classifications have been mapped and presented for consultation, reflecting an evolution in the research teams thinking:
- Unmodified Statistics NZ UAC
- Version 1 which includes some of the modifications listed below
- Version 2 which includes all the modifications listed below.
The team recognises that none of three classifications presented will be a ‘best fit’ and further versions will be developed and mapped in response to the feedback received in the co-design process.
The UAC is a generic classification that was not developed specifically for use in health research and policy and a number of significant modifications are being proposed to the UAC in order to generate a ‘fit for purpose’ GCH. Firstly, the connections between each urban area and its surrounding peri-urban zone will be recognised, and categorised appropriately. This is because people living on the outskirts of urban areas normally utilise the same health services (and have similar access to them) as residents in the nearby urban area. A 25 minute drive time is used to delineate the outer limits of this peri-urban zone. This 25 minute threshold aligns with one of the thresholds used in the UAC. Another major reason 25 minutes was chosen is because it is close to the threshold in the MoH PHO Services Agreement used to differentiate between urban and rural general practices (i.e. more than 30 minutes travel time from the GP to the nearest base hospital). This is commonly known as the ‘in-out rule’ by those working in the rural health sector and has formed the basis for much of the Rural Service Alliance Teams decisions on rural primary care funding allocation.
The second threshold proposed in the GCH is 60 minute drive time. This also corresponds to thresholds in the UAC and in the PHO Services agreement (PHOs are required to ensure 95% of their enrolled population have after-hours access to urgent care). It also aligns with the ‘golden hour’ of trauma care. A third drive time threshold is used to delimit the GCH categories R2 and R3. This recognises that some communities in NZ live a long distance from comprehensive healthcare. In Version 1 this is 120 minutes and in Version 2 it is 90 minutes.
It is intended that each of the GCH categories approximate to a different level of locally available healthcare with the need to travel to other areas in order to access more complex and specialist care. U1 areas contain ‘tertiary’ referral hospitals; U2 areas contain base hospitals that provide general medical, surgical, orthopaedic, obstetric, anaesthetic, emergency medicine and paediatric services, but limited subspecialty services; R1 areas would be expected to have locally available comprehensive 24/7 primary care +/- a rural hospital; R2 areas would be expected to contain some local primary care services, though these would not normally be comprehensive or available on a 24/7 basis. R3 residents would normally have to travel to access any healthcare services.
Several possible anomalies are evident in the proposed GCH. Timaru, Blenheim, Whakatane, and Masterton all provide base hospital level services for large geographic areas (Timaru and Masterton for entire DHB regions). One solution is to apply the U2 categorisation, despite them having an urban core with a population that is less than 30,000. Likewise Greymouth has a population less then 10,000 but serves a large geographic area. This anomaly has been recognised by the UAC which considers Greymouth a medium urban area. The GCH will do the same.
The level of locally available health services has been considered when constructing the GCH but health services are not an input variable in the way that populations size, density and travel time are. If health services access were included as an input variable this would compromise separate research aiming to understand the distribution of health services in rural and urban areas, since health service accessibility would be included as an ‘input’ and ‘output’ variable resulting in self-correlation.
The GCH aims to group together those communities that would be expected to have similar health service requirements based their population size, density, and travel time to other centres, using the matrix outlined below. It is important to remember this when considering the Versions presented.
The necessary building blocks of the UAC and SA1s (NZ Statistics Areas 1). SA1s normally contain 100 to 200 residents. Very sparsely populated SAIs cover large geographic areas. On occasions one boundary of a SA1 may be much more rural / remote than other parts of the SA1. This generates apparent anomalies when viewing maps of rural classifications built with SAIs and needs to be considered when viewing the maps. But because the methodology considers the population adjusted centroid of the SAI the actual number of individuals affected will very small.
The categories of the three classifications (1 to 3 above) that have been mapped are presented in table form in the blog.