healthmedicinet

Introduction

Type 2 diabetes continues to be a major health burden globally 1]. The changes in lifestyle and in particular the increasing rates of obesity are affecting
the increase of diabetes prevalence across the world 1]–3]. The prevention of diabetes is important but so is well-balanced diabetes care. Good
management of type 2 diabetes improves the quality of life of the patients, reduces
complications among patients 4], decreases the risk of comorbidities 5] and reduces the economic burden 6], 7].

Socio-economic inequalities in diabetes care do exist 8]. Achievement of the treatments targets in the diabetes care are affected by individual
9], 10] and neighborhood 11]–13] socio-economic status (SES). It is commonly believed that poor geographical accessibility
to health care services may lead to delayed care and underuse of health care and this
is believed to be the case especially among residents living in rural areas 14]. However, it should be kept in mind that rural health and health in general are interrelated
with broader social, economic, political, cultural, historical 15], 16] and spatial structures.

In Finland, primary care is available to all residents and is delivered mainly in
public health care centers by general practitioners (GPs). Most of the population
lives reasonably close to the nearest health service provider, but in rural areas
there are some long distances. In some areas, the availability of public transport
is inadequate. However, some of the chronic disease patients are entitled to reimbursements
for transportation to be able to attend the regular check-ups.

The aim of the study is to find out whether two different area classifications give
significantly different area-level results of achieving the targets of control and
treatment among type 2 diabetes patients exemplified by a Finnish region of North
Karelia, equivalent in area to New Jersey or 7/10 of Belgium. The focus is to reveal
and compare the possible spatial health care divergences by using 2-class (less detailed)
and 7-class (detailed) grid based classifications of urban and rural areas. The first
hypothesis is that the 7-class classification is better for showing differences in
urban and rural areas in the care of type 2 diabetes patients. The second hypothesis
claims that the longer the distance to the health center is the more it deteriorates
the achievement of control and treatment targets. The study exploits individual geo-referenced
type 2 diabetes patient record data from a regional primary health care patient database
combined with postal code area-level socio-economic variables, digital road data and
grid based classifications of areas.

Classifications of urban and rural areas

The absence of a generally accepted definition of urban and rural area types makes
it difficult to examine spatial health and health care inequalities in a valid way
and in particular to compare the results between different countries 17], 18]. This might be one reason for varying results on health and health care inequalities
in and between urban and rural areas. It has been suggested that these inequalities
should be examined across different settlement types and should not just rely on an
urban–rural dichotomy 16], 19], 20].

Often the definitions of urban and rural are based on population density and distance
to certain functions such as services 21], 22]. Definitions are usually developed for a certain purpose and generalization can lead
to lack of explanatory variation 22]. Traditionally, countries provide the classifications of urban and rural areas based
on different indicators, and usually these areas are consistent with the administrative
borders such as counties, municipalities, census blocks or census tracts.

However, grid based classifications also exist. In England and Wales urban and rural
areas (RUC: Rural–Urban Definition for Small Area Geographies) have been classified
for policy purposes by using 1 hectare grid cells 23]. The Organization for Economic Cooperation and Development (OECD) and the European
Commission have developed a grid based harmonized definition of cities in Europe which
improves cross-country analysis of cities 24]. Grid based and comparable on-task tailored classifications absorb more variation
than conventional classifications and thus could be more useful in health related
studies.

Helminen et al. 25] have developed a grid based 7-class classification of urban and rural areas for Finland
in 2014. Before that, the multiclass classifications of urban and rural areas for
various policy purposes were based on municipal borders. Old classification methods
became problematic and outdated as many municipalities merged in 2009–2012 creating
commuter belts where both urban and rural characteristics could be identified. The
new classification procedure is well documented and could be produced for other countries
as well.

The 2014 Finnish area classification divides urban areas into three (inner, outer,
peri-urban) classes and rural areas into four (local centers in rural areas, rural
areas close to urban areas, rural heartland areas and sparsely populated rural areas)
classes 26]. It depicts settlement structures focusing on population density, relative location,
land use and economic structures. This classification system uses geospatial data
represented by a 250 × 250 m grid of cells. Data on population, labor, commuting,
buildings, roads and land use have been used. Based on the data, variables describing
the amount, density, efficiency, accessibility, intensity, versatility and orientation
of the areas have been calculated. Each cell is classified into one of the seven classes
according to the defined criteria. All seven area classes are found in the study region
described later (Fig. 1).

Fig. 1. The study region of North Karelia, Finland. The area classifications used in the analyses:
the 2-class classification of population centers versus rural areas and the 7-class
classification of urban and rural area classes

The Finnish Environment Institute maintains a classification on population centers
(known also as statistical locality), which is provided by Statistics Finland. All
clusters of buildings with at least 200 inhabitants are defined as population centers
27]. The definition utilizes the building and population data of Statistics Finland’s
250 × 250 m grid data. The definition takes into account the population size, number
of buildings and their floor area. The distance between buildings included in population
centers is 200 m at maximum with certain exceptions. This categorization was also
used in this study to include a simple urban–rural dichotomy (Fig. 1). The patients living in population centers reside in urban areas and the patients
living outside the population centers reside in rural areas. The study region of North
Karelia is more rural as 70.3 % of the population lived in population centers compared
with the Finnish total urban population of 83.7 % in 2012 28].

Accessibility of health care services

The accessibility of health care services is affected by the locations of both the
health care provider and the patient. According to Penchansky and Thomas 29] accessibility (distance, transportation, travel time and cost) is one of the five
dimensions of access among availability (the supply of services), accommodation (hours
of operation, waiting times), affordability (price of services) and acceptability
(clients’ satisfaction). The poorer the accessibility is the larger the disadvantage
is made up by the friction of distance. Further on, accessibility can mean either
the potential or revealed accessibility 30]–32]. Potential accessibility consists of estimated values often based on surrogate variables,
whereas revealed accessibility means the actual use of health care services, thus
the health care service utilization 30]. Much of the research is focused on the methodology of measuring potential accessibility
31], 33]–37] but less on revealed accessibility and its effects on the outcomes of care.

Transportation options, transportation costs and distance to a health care provider
differ depending on the domicile of each patient. Poorer accessibility to health care
services is believed to lead to poorer health outcomes 14]. Commonly it is thought that utilization decreases as distance increases. However,
the effects of distance vary depending on the health service under consideration 30], 32].

Although distance may influence health care utilization, it is not a barrier to chronic
care 38], and patients will travel longer distances for check-ups or for chronic conditions
39]. Mixed evidence is found in the care of diabetes. Driving distance has not been associated
with care outcomes within urban settings in Canada 40], and differences in care outcomes have not been found between rural and urban patients
in Australia and the USA 41], 42]. In some cases driving distance has been associated with care outcomes in rural areas
in the USA 43], 44]. These diabetes related studies used very different definitions of urban and rural
or did not define them at all.