Areas in close proximity with similar values produce a spatial pattern indicative of positive spatial autocorrelation. Identifying groups of areas in close proximity to one another with high and low values is often of particular interest, suggesting a cluster of elevated risk with perhaps a common source. Many diseases are spatially constrained. Disease clustering is an important issue within public health. Prior researches have produced an incomplete and often counterintuitive picture. A framework has been developed to better understand how CDs is spatially clustered across governorates of Iraq.
Many studies in several countries were shown the prevalence of CDs is very
high and there is variation among studies. For example, in U.S. Wilper
et al. (2008) reported that 31.3% of uninsured and 43.5% of insured
had at least one of the six chronic illnesses. Furthermore, nearly 1 in 5 non-elderly
adults 17.5% had a history of at least 2 chronic conditions, whereas 6.0% had
3 or more. In Southeast Michigan, Jamil et al. (2009)
reported that prevalence of having any CD was 44% and the prevalence for Chaldeans
was 32%, for Arabs 44% and for Whites and Blacks 50% for each group. In Vietnam,
Minh et al. (2008) reported that 39% of the respondents
had at least one of the studied CDs. A Swedish study by Guez
et al. (2002) reported prevalence 43%. It was reported from HSE
(2008) that the prevalence of CDs was 6.5% in men and 4.0% in women. Of
402 Tunisians diagnosed with leukemia, 85.6 and 14.4% had chronic leukemia (Haouas
et al., 2011). In Sulaimani governorate, Iraq, Al-Windi
(2011) stated that more than halve population reported having CD. He found
the prevalence of CD higher among females 59.4% as compared to 49.4% in males
based on his questionnaire data conducted during May, 2006. Hystad
and Carpiano (2012) concluded that community-belonging was strongly related
to health-behaviour change in Canada and may be an important component of population
health prevention strategies.
In their study in Tilwana village and city of Elmenofia Governorate, Egypt,
Ali et al. (2010) concluded that pediatric asthma
is considered the most common chronic pediatric chest problem which has its
impact on a childs quality of life. The study conducted in U.S. by Liese
et al. (2012) suggested that neighbourhood characteristics related
to greater affluence, occupation and education are associated with higher type
1 diabetes risk based on multivariate generalized linear mixed models. Magalhaes
and Clements (2011) estimated geographical risk profile of anemia accounting
for malnutrition, malaria and helminthic infections in Burkina Faso, Ghana and
Mali (2003-2006). They found malnutrition and parasites make to anemia.
To understand the linkage between socioeconomic health indicators, investigation
should focus on features of areas rather than on compositional characteristics
of residents of the area which cannot fully describe the social environment
in which people live (Macintyre et al., 1993).
In recent years, a growing interest has been seen in examining the existence
of spatial autocorrelation in CD. However, governorates are tightly linked by
migration, commuting and inter-governorate trade. These types of spatial interaction
are exposed to frictional effects of distance, possibly causing the spatial
dependence of governorate conditions. So, the aim of the research is to study
spatial autocorrelation and geographical mapping regarding CDs.
Spatial autocorrelation is the term used for the interdependence of lattice
data over space. However, it was argued that lattice data are spatially correlated,
where Exploratory Spatial Data Analysis (ESDA) was used. The ESDA quantifies
the spatial pattern in order to increase the analysts knowledge of the
spatial system. As well as mapping plays an important role in monitoring health
status of people. Maps can reveal spatial patterns that is neither recognized
previously nor suspected from the examination of statistical tables. It reveals
high risk communities or problem areas (Lawson and Williams,
2001). Findings are expected to enhance health status monitoring and policing
interventions across governorates in Iraq.
Reducing inequality in CDs overall is not a primary objective but emergent
prosperity. The importance of this research objective emanates the hypothesis
stated that CDs can reduce people opportunities in keeping their job or/and
getting a high salary job. Also, to authors knowledge no studies used
spatial analysis techniques and geographical mapping in studying inequality
in CDs in Iraq. The importance of mapping was stated by Koch
(2005): why make the map if detailed statistical tables carry the same results?
Perhaps the most important reason for studying spatial statistics is not only
interested in answering the "how much" question but the "how much is where"
question (Schabenberger and Gotway, 2005). Often, this
question is related to the need for public health authorities to monitor unusual
aggregations of diseases in localized areas within their area of authority.
These concerns may be routine in that there may be a need to provide surveillance
of particular diseases and to be aware of any atypical geographical distributions.
Therefore, the usefulness of the paper is to suggest where to intervene geographically.
The main purpose of this study is to measure spatial clustering in CDs in Iraq,
with a special emphasis on geographic distribution. In light of these: (1) The
existence of spatial global clustering and local clusters for CDs were investigated
and (2) Mapping was displayed for CDs and its local Morans Ii
values. The study design is a cross-sectional analysis in a census survey conducted
in Iraq in 2007.
MATERIALS AND METHODS
Data: Data were collected from the tabulation report of Iraq household socio-economic survey conducted in 2007. For each of (N = 18) governorates, transformed CDs data were applied. CDs includes diabetes, high blood pressure, chronic inflammation, cancer, psychiatric, nervous and sensory, cardiac, respiratory, digestive, kidney disease, anemia and other. With respect to sex, CDs for males 11.5% was slightly lower than CDs for females 14.6%. However, these percentages vary geographically. In Kurdistan region, Baghdad and other governorates were 19.5, 14.9 and 10.9%.
Analysis: Data analysis involved five steps. In step 1, CDs variable was tested for normal distribution. It wasnt found to follow normal distribution. Therefore, CDs was transformed to follow normal distribution using LISREL software. LISREL scales normal scores so that transformed variable has the same sample mean and standard deviation as the original variable. Thus, normal score is a monotonic transformation of original score with same mean and standard deviation (this characteristic can be considered as an advantage in this transformation) but with the values of skewness and kurtosis much reduced. In step 2, visual inspection based on quantified gradients for CDs was conducted using quartiles. Step 3 included the calculation of global Morans I-statistic for CDs to detect global clustering and the significance of I-statistic using permutation test was examined. Step 4 involved the calculation of local Morans Ii for ith governorate and its p-value using Monte Carlo simulation to detect local clusters. In step 5, visual inspection for local Moran values was inspected based on choropleth mapping.
|| Study area shows all governorates with their ID and the neighbours
of each governorate
Maps showing diseases can help elucidate the cause of disease and may be used to generate hypotheses of disease causation. CD variable was categorized by four intervals. These intervals were used for all maps using darker shade of gray to indicate increasing value. Such approach enables qualitative evaluation of spatial pattern. In neighbourhood researches, neighbours may be defined as areas border each other or within a certain distance of each other. In this research neighbouring structure was defined as governorates shares a boundary. The second order method (queen pattern) which included both the first-order neighbours (rook pattern) and those diagonally linked (bishop pattern) was used. A neighbourhood structure of Iraqs governorates is explained in Fig. 1, where identification numbers (ID) of neighbours for each governorate are shown.
To construct a choropleth map, data for enumeration governorates are typically grouped into classes and a gray tone was assigned to each class. Although, maps allow visual assessment for spatial pattern, they have two important limitations: their interpretation varies from person to person and there is possibility that a perceived pattern is actually the result of randomness and thus not meaningful. For these reasons, it makes sense to compute a numerical measure of spatial pattern which can be accomplished using spatial autocorrelation.
Identification of global clustering: The goal of a global index of spatial autocorrelation is to summarize the degree to which similar observations tend to occur near to each other in geographic space. In this exploratory spatial analysis, spatial autocorrelation using standard normal deviate (z-statistic) of Morans I under normal assumption was tested. Morans I coefficient is used to measure the strength of spatial autocorrelation in regional data. The null hypothesis of no spatial autocorrelation or spatially independent versus the alternative of positive spatial autocorrelation is as follows:
H0: No clustering exists (no spatial autocorrelation)
H1: Clustering exists (positive spatial autocorrelation)
Morans I is calculated as follows (Cliff and Ord, 1981):
where, N is 18 is the number of governorates, wij is 1 is a weight denoting the strength of the connection between two governorates i and j, otherwise, wij is zero and xi and xj represents CDs in the ith and jth governorate, respectively.
To test the significance of global Morans I, z-statistic which follows
a standard normal distribution was applied. It is calculated as follows (Weeks,
Permutation test was applied. A permutation test tells us that a certain pattern in data is or is not likely to have arisen by chance. The observations of CDs were randomly reallocated 1000 times with 1000 of spatial autocorrelations were calculated in each time to test the null hypothesis of randomness. The hypothesis under investigation suggests that there will be a tendency for a certain type of spatial pattern to appear in data. Whereas the null hypothesis says that if this pattern is present, then this is a pure chance effect of observations in a random order.
The analysis suggests an evidence of clustering if the result of global test is found significant; though it doesnt identify the locations of any particular clusters. Besides clustering that represents global characteristic of CDs, the existence and location of localized spatial clusters are of interest in geographic sociology. Accordingly, local spatial statistic was advocated for identifying and assessing potential clusters.
Identification of local clusters: A global index can suggest clustering
but cannot identify individual clusters (Waller and
Gotway, 2004). Anselin (1995) proposed local Morans
Ii statistic to test local autocorrelation. Local spatial clusters,
sometimes referred to as hot spots, may be identified as those locations or
sets of contiguous locations for which the local Morans Ii
is significant. However, Morans Ii for ith governorate may
be defined by Waller and Gotway as:
where, analogous to global Morans I, xi and xj represents CDs in the ith and jth governorate, respectively, Ni = number of neighbours for ith governorate and S is the standard deviation. It is noteworthy to mention that the number of neighbours for ith governorate were taken into account by the amount:
where, wij was measured in the same manner as in Morans I statistic. Local Moran statistic was used to test the null hypothesis of no clusters.
The most basic definition of cluster is any area within the study region of significant elevated risk. Cluster could be due to either aggregation of high values, aggregation of low values, or aggregation of moderate values. Thereby, high values of Ii suggesting a cluster of similar (but not necessarily large) values across several governorates and low value of Ii suggesting an outlying cluster in a single governorate i (being different from most or all of its neighbours). A positive local Moran value indicates local stability; such as governorate has high/low value surrounded by governorate has high/low value. A negative local Moran value indicates local instability, such as governorate has low value surrounded by governorate has high value or vice versa. However, each governorates Ii value was mapped to provide insight into the location of governorates with comparatively high or low local association with their neighbouring values. In statistical analysis, all programs were performed using SPLUS8 Software.
Descriptive statistics were calculated for transformed CDs. The mean and standard deviation were found 12.72 and 4.08, respectively. Skewness and kurtosis were found 0.00 and -0.10, respectively. The five-number summary of CDs data set consisted of minimum, maximum and quartiles written in increasing order: Min = 4.64, Q1 = 9.83, Q2 = 12.72, Q3 = 15.62 and Max = 20.81. From the five-number summary, the variations of the four quarters were found 5.19, 2.89, 2.90 and 5.19, respectively, where the 1st and 4th quarters have the greatest variation of all.
Figure 2a and b show visual insight for
CDs and its local Moran values, respectively. Darkest shade corresponds to the
highest quartile. These maps display geographical inequalities across governorates
of Iraq. Based on visual inspection taken from Fig. 2a, an
overall worsening pattern (higher scores) was found in eastern-northern and
western-southern parts. The same was approximately found regarding visual inspection
taken from Fig. 2b for local Moran values of CDs that worsening
pattern was found in eastern-northern and western parts.
Suggestion of spatial clustering for CDs that follows a visual inspection of
mapping was not confirmed by global Morans I of .06 with an associated
z-statistic of 0.91 and p = 0.365. Although, global clustering was not confirmed,
permutation test was done.
||Choropleth maps show visual insight for: (a) CDs variable
and (b) its local Moran values
||Shows CDs variable, its transformed, its local
Morans Ii value and its corresponding p-value
|Bold values are significant at 0.10 level
Permutation p = 0.175 wasnt found significant. Thus, the null hypothesis
of no spatial autocorrelation wasnt rejected accordingly. But, three significant
local clusters (hot spots) were found. These clusters: 1, 5 and 7 are located,
as shown in Fig. 2b, in northern and western parts. Table
1 shows local Moran values of CDs variable and their p-values.
Simulated data are useful for validating the results of spatial analysis. However,
using Monte Carlo simulation, 9 999 random samples, eighteen values for each
sample, were simulated. The process of simulation was conducted under standard
normal distribution to calculate p-values for local Moran values of CD. When
the word simulation is used, it is referred to an analytical method meant to
imitate a real-life system, especially when other analyses are too mathematically
complex or too difficult to reproduce. While research results were specific
to these data, the case study helps to identify general concepts for future
The above framework has revealed some noteworthy findings. Such findings allow policy makers to better identify what types of resources are needed and precisely where they should be employed. This study adds to the global body of knowledge on the utilization of spatial analysis to strengthen the research-policy interface in developing countries such as Iraq. Although, this work was conducted as part of a wider study, its immediate implications are more for policy makers and practitioners than for researchers. Spatial patterns had shown concentration of highly CDs in old governorates such as Duhouk, Erbil and Alanbar. The contribution to literature on CDs was shown by explaining where these patterns were located in Iraq. Findings of this study suggest that tackling CDs is a high priority. It should foster efforts to ensure that CDs prevention strategies include urban poor governorates.
Maps provide powerful means to communicate data to others. Unlike information
displayed in graphs, tables and charts; maps also provide bookmarks for memories.
In this way, maps are not passive mechanism for presenting information. Usually,
in the spatial analysis and geographical mapping, small spatial areas should
be used such as districts, counties
etc. But, in this research governorates
were considered somewhat larger than for example districts due to data were
not available for smaller areas. Most often the word neighbourhood
suggests a relatively small area surrounding individuals homes. But researchers
commonly make use of larger spatial area such as census tracts (Coulton
et al., 2001). Often, choices about neighbourhood spatial definitions
were made with respect to convenience and availability of contextual data rather
than study purpose (Schaefer-McDaniel et al., 2010).
Schaefer-McDaniel et al. (2010) stated that researchers
might utilize census data and thus rely on census-imposed boundaries to define
neighbourhoods even though these spatial areas may not be the best geographic
units for the study topic.
As noted by Waller and Jacques (1995), the test for
spatial pattern employs alternative hypotheses of two types; the omnibus not
the null hypothesis or more specific alternatives. Tests with specific alternatives
include focused tests that are sensitive to monotonically decreasing risk as
distance from a putative exposure source (the focus) increases. Acceptance of
either types (the omnibus or a more specific alternatives) only demonstrates
that some spatial pattern exist and does not implicate a cause (Jacques,
2004). Hence, the existence of a spatial pattern alone cannot demonstrate
nor prove a causal mechanism.
It is very well known that employment is considered a major source for household
income (HI). Generally, low income leads to several CDs. Iraq before, in and
after 2007 faced specific challenges with regard to jobs in its state-owned
enterprises. With over 500,000 workers on the payroll, state-owned enterprises
are a major source of employment. Analysis of other episodes of conflict in
Iraq indicated a very strong reciprocal relationship between the lack of security
and high unemployment (World Bank, 2006). While, reconstruction
and associated public-sector jobs are important in the initial phase of Iraqs
recovery, they will not create a sufficient number of jobs to meet the populations
needs in the long term; even if recovery is on a massive scale. Most researchers
recommend designing global development strategies that focus on job creation
and income generation and incorporate elements of basic social protection and
social dialogue at the global and local levels as an attempt to reduce inequalities
The application of statistical techniques to spatial data faces an important
challenge, as expressed in the first law of geography: everything is related
to everything else but closer things are more related than distant things
(Tobler, 1979). The quantitative expression of this principal
is the effect of spatial dependence, i.e., when observed values are spatially
clustered, the samples are not independent. Increasing in CDs level in governorate
i generates increasing in CDs level in governorate j. This mechanism of transmission
leads to spatial autocorrelation in CDs.
Anselin (1995) stated that indication of local patterns
of spatial association may be in line with a global indication, although this
is not necessarily be the case. It is quite possible that local pattern is an
aberration that global indicator would not pick up, or it may be that a few
local patterns run in the opposite direction of global spatial trend. Local
values that are very different from the mean (or median) would indicate locations
that contribute more than their expected share to the global statistic. These
may be outliers or high leverage points and thus would invite closer scouting.
However, this is found in this research. Although, global Moran statistic was
not found significant, three local Moran values were found significant. Limitation
of this study was that causal relationship cant be drawn due to this can
only be done in prospective studies.
The obvious question after finding significant clusters in CDs is-why? Could
this pattern associated by the spatial pattern of socioeconomic indicators such
as HI or by the limitation of economic resources? However, further research
is required regarding this bivariate spatial association between CDs and other
socioeconomic indicators. The lack of adequate health care, health care system
and adequate management could also belong to the reasons for poor health. This
research will be our interest in Iraq and other developing countries in the
near future. As stated by Barkera (2012), prevention
of chronic disease and an increase in healthy aging require improvement in the
nutrition of girls and young women. Many babies in the womb in the Western world
today are receiving unbalanced and inadequate diets. Many babies in the developing
world are malnourished because their mothers are chronically malnourished.
In essence, it should be emphasized that CDs problem cannot overcome in the
short-run; but long-term efforts are needed to tackle inequalities across governorates.
In turn, enabling the economy to create more job opportunities and establish
new projects, especially in governorates that are found as hot spots. It means
that the place of the problem is now clearly shown. Health inequalities are
ubiquitous around the world. Thus, fresh perspectives to tackle inequalities
are always welcomed by the research community invested in reducing and eventually
eliminating these inequalities. Finally, this kind of studies should be conducted
periodically in light of the changing of socioeconomic and political conditions.
Conclusions are comprehensive in at least four aspects. First, visual inspection shows that high level in CDs was concentrated in eastern-northern and western-southern governorates. While low level in CDs was concentrated in middle and some northern and southern governorates. Second, several governorates such as Alanbar, Kerbela and Babil are not observed visually as hot spots. But, after considering the information of their neighbours, the pattern of their hot spots can be obviously seen. Third, global clustering in CDs was not found significant. However, three governorates: Duhouk, Erbil and Alanbar were found significant local clusters. Forth, from negative local Moran values, looking at the local variation, some governorates were seen as areas of dissimilarity, such as Baghdad and Alnajaf. These conclusions are useful for organization and follow-up of medical care. By providing the necessary knowledge, patients and family are encouraged to effectively manage the disease process and improve their quality of life.
The analytical approach used here delineates governorates of relatively high CDs. This permits policy makers to develop strategies that minimize this inequality. Policy which pays attention to area characteristics will probably reduce CDs inequality. Consequently, this will generally improve population health in affected governorates. In summary, this study supports the hypothesis of spatial clustering in CDs at governorate level that probably reflects inequality distribution in several socioeconomic indicators. Further research is required to investigate spatial clustering in several socioeconomic indicators attempting to explain hot spots in CDs.