Content uploaded by Guofa Zhou
Author content
All content in this area was uploaded by Guofa Zhou
Content may be subject to copyright.
SPATIO-TEMPORAL DISTRIBUTION OF PLASMODIUM FALCIPARUM AND
P. VIVAX MALARIA IN THAILAND
GUOFA ZHOU, JEERAPHAT SIRICHAISINTHOP, JETSUMON SATTABONGKOT, JAMES JONES,
OTTAR N. BJØRNSTAD, GUIYUN YAN,
AND LIWANG CUI
Department of Biological Sciences, State University of New York, Buffalo, New York; Department of Entomology, Pennsylvania
State University, University Park, Pennsylvania; Vector Borne Disease Training Center, Ministry of Public Health, Saraburi,
Thailand; Department of Entomology, Armed Forces Institute of Medical Sciences, United States Army Military Component,
Bangkok, Thailand
Abstract. Malaria incidence data at the district level from 1997 to 2002 and total malaria case data from 1965 to 2002
in Thailand were analyzed to determine the spatial and temporal dynamics of Plasmodium falciparum and P. vivax
malaria incidence. Over the 37-year period, there was a 35-fold reduction in the incidence rates of P. falciparum malaria
(11.86% in 1965 versus 0.34% in 2002) and a 7-fold reduction in P. vivax malaria (2.89% in 1965 versus 0.40% in 2002).
The incidence ratio of P. falciparum to P. vivax malaria was reduced from 4.1 to 0.8 during this period. Malaria incidence
rate exhibited the most rapid reduction between 1975 and 1985, coinciding with the introduction of a combination of
antifolate drugs (sulfadoxine-pyrimethamine). The distribution maps of P. falciparum and P. vivax malaria incidence
rates indicated a high spatial heterogeneity. The Thailand-Myanmar and Thailand-Cambodia border areas, where
migration of foreign workers was pronounced, had the highest incidence rates for P. falciparum, P. vivax, and mixed-
species infections. Transition probability analysis based on the malaria incidence rate among Thai residents indicated
that there was an overall trend of decrease in the number of malaria cases and the number of high incidence districts
between 1997 and 2002. High spatial variation in malaria incidence and local human migration patterns suggest that
malaria control measures need to be adjusted according to local environmental and demographic settings.
INTRODUCTION
Malaria is a major public health problem in Thailand.
1,2
Annual reported malaria cases in Thailand have continued to
decrease over the past two decades and have disappeared
from most of the major cities.
2
However, people in rural ar-
eas, especially in villages on the Thailand-Myanmar and Thai-
land-Cambodia borders and forested mountain areas, remain
at great risk.
3–6
In these endemic areas, malaria transmission
has been considered to have a close association with the forest
and movement of the human population.
1,2,7,8
In particular,
the seasonal migration of cross-border laborers has been sus-
pected as a leading cause of malaria transmission in these
areas.
2,9
However, little is known about spatial patterns and
dynamics of malaria in Thailand.
In Thailand, all four species of human malaria occur, but
the vast majority of malaria cases are caused by Plasmodium
falciparum and P. vivax infections. Since the appearance of
chloroquine (CQ)–resistant P. falciparum strains in the 1960s,
multidrug-resistant P. falciparum has emerged and spread to
most areas of this country.
4,10–12
In contrast, drug resistance
in P. vivax has not been reported in Thailand.
13
Because ma-
laria caused by these two species is treated with different drug
regimens, correct diagnosis of malaria species is essential. The
existence of different malaria species presents opportunities
for mixed-species infections and cross-species interactions,
which could affect the outcome of the disease. Previous cross-
sectional studies have provided indirect evidence of cross-
species interactions. For example, studies on the Pacific island
of Vanuatu found that P. falciparum malaria was predomi-
nant in the long wet season and P. vivax malaria in the dry
season.
14,15
More direct evidence of cross-species interactions
between P. falciparum and P. vivax was obtained from a well-
designed study in a hyperendemic area of Papua New Guinea,
where nonindependent and sequential episodes of P. falci-
parum and P. vivax malaria appear to be regulated by para-
site density.
16
In Thailand, a clinical follow-up study found
that P. vivax malaria appeared 15–65 days after the treatment
of P. falciparum malaria, suggesting an effect of inter-species
interactions.
17
However, large-scale, cross-sectional analyses
of the relationship between the two malaria species have not
been performed.
The history of malaria control strongly emphasizes the
implementation of integrated approaches. Geographically
based monitoring systems and spatial statistics are an impor-
tant component of the contemporary malaria integrated man-
agement system.
18,19
This novel approach requires the devel-
opment of models that can be used to monitor and predict
large-scale spatio-temporal dynamics of malaria. In this study,
we analyzed malaria incidence data on different spatial scales
and the association between P. falciparum and P. vivax ma-
laria. Understanding these spatial processes will help develop
statistical models of malaria endemicity, which, in turn, will
guide our decisions in malaria case management.
MATERIALS AND METHODS
Malaria incidence data. The Malaria Control Program of
the Department of Communicable Disease Control, Ministry
of Public Health of Thailand, was responsible for monitoring
and implementing the malaria control programs and archiving
the malaria incidence data. Malaria incidence refers to the
number of symptomatic clinical malaria episodes (confirmed
by microscopy) occurring in the population within a specific
year. This study analyzed countrywide malaria incidence dy-
namics from 1965 to 2002. District-level malaria incidence
data covering 926 districts from 1997 to 2002 were compiled
from the records at district malaria clinics and hospitals. The
recorded number of foreigner malaria cases at the district
level was collected from 1990 to 2001. A more detailed analy-
sis was focused on the Mae Sod District in Tak Province
where monthly malaria case data were available for three
malaria clinics from 2000 to 2002. These clinics, located in
Mae Sod, Mae Kasa, and Mae Kuedlong, are approximately
10 km apart. This dataset was further categorized by the ori-
gin of the patients (local residents versus foreigners). For all
Am. J. Trop. Med. Hyg., 72(3), 2005, pp. 256–262
Copyright © 2005 by The American Society of Tropical Medicine and Hygiene
256
clinical data, patient information was excluded. Malaria cases
from the two most prevalent parasite species (P. falciparum
and P. vivax) were further divided and used for the subse-
quent temporal and spatial analyses.
Historical demographic data of Thailand for the period of
1965–2002 were obtained from the National Statistics Office
of Thailand and Asia-Pacific Economic Committee.
20,21
Countrywide malaria incidence rate was calculated based on
malaria incidence and population data.
22
The district-level
population census of 2000 was used to generate the map of
malaria incidence rate (number of cases per 100 population)
for the year 2000. The population growth rate from 1997 to
2002 at the provincial level was used to estimate district-level
population of years other than 2000, assuming that all districts
in a given province had the same population growth rate.
20
At
the provincial level, the annual total malaria cases were sum-
marized and malaria incidence rates were calculated. Clinic-
level malaria incidence rate was calculated using the number
of malaria cases at each malaria clinic in Mae Sod district and
population data of 2003 at each designated town where the
clinic is located.
Spatio-temporal autocorrelation analysis. Spatial analysis is
a statistical technique for describing the spatial variations
exhibited by the response process of a given phenomenon.
23–25
A spatial autocorrelation statistic is used to quantify the de-
gree of association of a response variable with adjacent points
or areas. In this study, we used Moran’s I statistic to calculate
the autocorrelation of malaria incidence rates for sets of
points (centroid of each district) that are spatially adjacent in
different distance classes.
23
Moran’s index I is defined as
I =
n
S
0
兺
i
兺
j⫽i
w
ij
共
x
i
− x
兲
共
x
j
− x
兲
兺
i
(x
i
− x)
2
where x
i
is the malaria incidence rate of district i of a given
year and x¯ is the mean incidence rate for all districts, w
ij
is the
weight factor that defines the spatial relationship between
district i and district j, and
S
0
=
兺
i
兺
j⫽i
W
ij
.
The correlogram I(d) quantifies the correlation between lo-
cations separated by distance d.
26
This is constructed by set-
ting the weight w
ij
to 1 if the distance between districts i and
j falls within the tolerance interval [d−d
0
, d + d
0
], 0 other-
wise. Values of Moran’s I range from 1 to −1, with an ex-
pected value of −1/(n − 1), where n is the number of localities.
To calculate the distance matrix, the universal transverse
mercator coordinate of each district was generated using the
centroid coordinates method of ArcView.
27
Transition probability analysis. Transition probability mod-
els, which describe temporal transition of a variable from one
nominal value (or state) to another, have been used to char-
acterize outbreak patterns of infectious diseases from histori-
cal data.
28–30
In this study, we used the multistate Markov-
chain models to quantify the transition of districts from one
class to another.
31
We categorized the malaria incidence rates
into five classes: 0 cases (class 1),>0to<0.1cases (class 2),
ⱖ 0.1 to < 1 cases (class 3), ⱖ 1 to 5 cases (class 4), and > 5
cases (class 5) per 100 population. The transition probability
was calculated based on the six-year district-level malaria in-
cidence data (1997–2002).
Statistical analysis. Correlation analysis was used to mea-
sure the correlation between incidence rates of P. falciparum
and P. vivax malaria at both the country and district levels.
32
The Student’s t-test was used to test differences between the
national average and province/district-level incidence rates,
between clinic-level P. falciparum and P. vivax malaria inci-
dence rates, and between clinic-level foreign and Thai malaria
case numbers. The chi-square test with Yates’ correction for
continuity was used to test the association between P. falci-
parum and P. vivax at the clinic level.
RESULTS
Trend of countrywide malaria incidence. Malaria incidence
rates in Thailand showed an overall tendency of decline over
the past four decades (Figure 1A). During this period, there
was a 35-fold and 7-fold reduction in the incidence rates of P.
falciparum malaria (11.86% in 1965 versus 0.34% in 2002)
and P. vivax malaria (2.89% in 1965 versus 0.40% in 2002),
respectively. Meanwhile, the incidence ratio of P. falciparum
malaria to P. vivax malaria was reduced from 4.1 to 0.8. The
most rapid reduction in malaria incidence rates occurred be-
tween 1975 and 1985. We further found a significant correla-
tion between the incidence rates of malaria caused by the two
most prevalent parasite species, P. falciparum and P. vivax
(R ⳱ 0.83, P < 0.001). Therefore, despite temporal fluctua-
tions, both species had similar trends of decline, although
annual P. falciparum malaria case numbers have decreased at
a faster pace since the early 1980s. It is interesting to note that
the fluctuations of malaria incidence rates, especially P. fal-
ciparum malaria, were closely associated with the changes of
antimalarial drugs imposed by the national drug policies of
Thailand (Figure 1).
4,13,33
In Thailand, increasing proportions of malaria cases were
from cross-border migratory foreign workers, and these cases
were especially concentrated in districts bordering Cambodia
and Myanmar. In the past 12 years, foreigner malaria case
numbers remained relatively constant (Figure 1B), suggesting
that cross-border seasonal labor may play an important role
in malaria transmission in Thailand. Indeed, the provinces
with the highest incidence rate border Myanmar and Cambo-
dia. Population movements in these areas, together with the
high drug pressure, were considered responsible for the de-
velopment and spread of mefloquine-resistant P. falciparum
in western Thailand.
34
District-level patterns of malaria incidences. At the district
level, P. falciparum and P. vivax malaria exhibited similar
spatial patterns and temporal dynamics. Analysis of data from
1997 to 2002 showed a mean correlation of 0.82 (range ⳱
0.77–0.88, P < 0.01) between the two species. Correlograms of
district malaria incidence rates showed significant spatial au-
tocorrelation up to 100 km (Figure 2), indicating that malaria
incidences in Thailand were not randomly distributed, but
rather occurred as clusters among adjacent districts. Despite
the overall decrease in malaria cases during this period, the
extent of spatial clustering, as measured by the spatial auto-
correlation, did not change dramatically.
Based on malaria incidence rates, we divided the districts of
Thailand into five classes (Figure 3). Class 1 represents ma-
laria-free areas, whereas class 5 is the area with highest inci-
SPATIO-TEMPORAL PATTERNS OF MALARIA INCIDENCE 257
dence rates. The areas in class 5 are located mainly along the
Thailand-Myanmar and Thailand-Cambodia borders (Figure
3). To test the persistence of the malaria status of each class,
we performed transition probability analysis. The results
demonstrated that class 1 districts had the highest probability
(85%) of remaining malaria free, whereas the most endemic
districts in class 5 had a 54% probability of remaining in this
class (Table 1). As for the probability of malaria incidence
rate reduction, districts in class 5 showed a 46% probability of
changing to lower classes. In general, the result of the tran-
sition matrix reflected an overall trend of decline in malaria
incidence rates in terms of the number of total cases and the
districts with the highest incidence rates (i.e., in classes 4 and
5) from 1997 to 2002.
FIGURE 1. A, Dynamics of malaria incidence rate (number of cases per 100 population) among Thai residents and antimalarial drug policy
changes in Thailand from 1965 to 2002 and B, malaria incidence among Thai residents and foreigners diagnosed and treated in Thailand from 1991
to 2002. CHL ⳱ chloroquine; SP ⳱ sulfadoxine-pyrimethamine; QT ⳱ quinine-tetracycline; MSP ⳱ mefloquine plus SP; M ⳱ mefloquine;
ATS ⳱ artemisinin.
ZHOU AND OTHERS258
Clinic-level malaria incidence. We have further focused our
analysis on three clinics in Mae Sod district in Tak Province.
In general, Tak Province had higher malaria incidence rates
(5.68 cases per 100 population) than the national average
(0.16 cases per 100 population) over the six-year period
(t ⳱ 9.36, degrees of freedom [df] ⳱ 5, P < 0.001). Data from
the three clinics in Mae Sod district showed that malaria in-
cidence rates in Mae Sod (0.67 cases per 100 population) and
Mae Kasa (0.32 cases per 100 population) were significantly
higher than the national average (P < 0.05 for both sites).
Despite overall high malaria incidence rates in Tak Province,
Mae Kuedlong had significantly lower rates (0.04 cases per
100 population) than the national average (P < 0.01). Similar
to the nationwide malaria trend, the proportions of P. vivax
malaria in 2002–2003 had increased and were significantly
higher than those of P. falciparum malaria. Furthermore,
when the relative incidence of each malaria species in these
three clinics were compared, we found that there were sig-
nificantly more P. vivax malaria cases in Mae Kasa (P. falci-
parum:P. vivax ⳱ 282:385, t ⳱ 3.21, df ⳱ 11, P < 0.01),
whereas the other two clinics did not show such a bias, sug-
gesting that malaria species composition varies on a micro-
geographic scale.
One unique feature of malaria epidemiology in Thailand is
that the numbers of malaria cases from foreigners have re-
mained relatively constant for the past 12 years. Located at
the Thailand-Myanmar border, Mae Sod district represents
an area where malaria case numbers from foreign travelers
were significantly higher than from local residents (malaria
cases in 2000–2001, Thai:Foreigner ⳱ 4,420:9,339, t ⳱ 7.02,
df ⳱ 23, P < 0.001). This indicates that cross-border popula-
tion movement may contribute tremendously to malaria
transmission in this border area. Interestingly, among the for-
eigner malaria cases, there were significantly more P. falci-
parum than P. vivax malaria cases, whereas in local residents
malaria cases from the two species were equally abundant.
Previously, we have shown that the clear seasonality of ma-
laria incidence in this area, with two peaks occurring in May–
July and October–November, is closely associated with the
patterns of rainfall.
35
When the malaria case data from for-
eigners and local residents were analyzed separately, the
seasonal peaks were most obvious for the foreigner malaria
cases. In contrast, malaria cases in local residents were
not markedly seasonal, and the case numbers in the peak
months were not significantly different from those in other
months.
In Mae Sod, all four human malaria species are present
with P. vivax and P. falciparum as the most prevalent species.
Using merozoite surface protein 1 (MSP-1) and MSP-2 poly-
morphism data, we have previously shown that mixed-strain
infection of P. vivax malaria in this area was as high as 35%
despite its low endemicity.
35
An earlier study in Trad Prov-
ince of Thailand detected 20% mixed-species infections.
36
However, analysis of the two-year data from three clinics in
Mae Sod did not show a significant association between P.
falciparum and P. vivax in either Thai or foreign cases (P <
0.001 for each species at each of the three clinics, by chi-
square test).
32
For example, less than 2.3% of all malaria
cases in Mae Sod district were microscopically diagnosed as
mixed-species infections. Such a decrease in mixed species
TABLE 1
Matrix of transition probabilities among the five malaria incidence
rate classes
Transition
probability
Transitioned class
12345
Initial class
1 0.85 0.14 0.00 0.00 0.00
2 0.18 0.77 0.05 0.01 0.00
3 0.01 0.24 0.68 0.07 0.01
4 0.00 0.01 0.33 0.60 0.06
5 0.00 0.00 0.00 0.46 0.54
FIGURE 3. Dynamics of spatial distribution of malaria incidence
rates (number of cases per 100 population) at the district level be-
tween 1997 and 2002. This figure appears in color at www.ajtmh.org.
FIGURE 2. Yearly spatial correlograms of malaria incidence rate
(cases per 100 population) for the period 1997–2002.
SPATIO-TEMPORAL PATTERNS OF MALARIA INCIDENCE 259
infection rate might be due to the overall reduction in malaria
transmission and/or the insensitivity of microscopic examina-
tion to detect mixed species infections.
DISCUSSION
In this study, we performed a retrospective analysis of ma-
laria incidence data for the past 37 years in Thailand. Al-
though a gradual decrease in nationwide malaria cases has
been observed over this period, there was high spatial het-
erogeneity in malaria incidence rates across the country.
High-incidence regions were concentrated near areas border-
ing Myanmar and Cambodia. Further analysis of malaria in-
cidence data in Mae Sod district showed great variation in
malaria incidence rates and species composition between lo-
cal and foreign patients. On the Thai side, political stability
has fostered a good public health infrastructure. In Mae Sod,
there are three government-funded malaria clinics offering
free diagnosis and treatment of malaria. On the Myanmar
side, such services are not present at this time. As a result,
many patients simply cross the border to seek free malaria
treatments. Therefore, the significantly higher foreigner case
numbers in Mae Sod may have actually resulted from in-
creased human migration and perhaps indicate higher levels
of malaria transmission in Myanmar. These foreign malaria
cases may have contributed to local malaria transmission,
since Thai residents in Tak Province had the highest malaria
incidence rates in the country. Given the possibility that mi-
gratory foreign workers are an important source of malaria
transmission, the change in the patterns of migration could
have impacted on the population at risk and the incidence of
reported cases. Unfortunately, accurate records on the num-
ber of migratory laborers and the patterns of migration are
not available. Nonetheless, the results of this retrospective
analysis suggest that control of malaria in these high-
incidence areas requires close monitoring of malaria in mi-
gratory foreigners.
Based on malaria incidence rates, 926 districts have been
categorized into five classes in our analysis. Spatial autocor-
relation analysis showed geographical association of districts
to form clusters, suggesting that malaria transmission in one
district is directly or indirectly associated with transmission in
neighboring districts. This may be due to regional similarities
in climatic and environmental factors that are linked to the
dynamics of vectors. Although this was not directly tested due
to the lack of environmental data, the malaria distribution
map showing the concentration of the disease in the forested
areas and areas bordering Myanmar and Cambodia favors
such a connection. In addition, the clustering of high-
incidence districts may be related to socioeconomic factors
that affect the effectiveness of vector control programs. An
earlier study has clearly shown that malaria is directly linked
to poverty in Thailand.
37
We have noted that in certain high-
incidence areas, malaria control programs include govern-
ment-funded free malaria clinics and local vector control ef-
forts (mostly application of residual insecticide). However,
the effectiveness of the vector control program on malaria
control has not been evaluated. State transition analysis
found that each malaria incidence class had a high probability
(54–85%) of remaining in the same class. On one hand, this
indicates that malaria will persist in Thailand, especially in the
rural and border areas, which further stresses the need for a
continued and strengthened malaria control program. On the
other hand, the results are also encouraging, showing signifi-
cant improvements of malaria status in the central regions
and chances for improvement in the bordering endemic re-
gions.
The Malaria Control Program of the Department of Com-
municable Disease Control is responsible for monitoring ma-
laria incidence and changes in the national antimalarial drug
policies. In the mid 1970s, malaria cases from both parasites
were on the increase because the drug used at that time, CQ,
became less effective. Consequently, a new combination of
antifolate drugs (sulfadoxine-pyrimethamine [SP]), was intro-
duced in 1975. This resulted in a sharp decrease in P. falci-
parum malaria, but P. vivax malaria was not greatly af-
fected.
13
Since 1980, the national drug policy has changed
four times in response to the drug resistance problem in
P. falciparum. To date, P. falciparum has developed resis-
tance to CQ, SP, and mefloquine in succession, and multidug-
resistant parasites have spread to most malarious areas of this
country.
38
To deal with multidrug resistance, a mefloquine-
artesunate combination became the standard treatment for
P. falciparum malaria in 1995. Overall, P. falciparum re-
sponded well to drug changes; each drug change resulted in
an immediate sharp decrease in P. falciparum malaria cases.
Since 1995, the P. falciparum malaria incidence rate contin-
ued to decrease each year, but the P. vivax malaria incidence
rate showed only a slight decline. As a result, P. vivax became
the most prevalent malaria parasite in Thailand at the turn of
the century. Such a difference in response to control measures
may be directly linked to the intrinsic biologic properties of P.
vivax, such as early gametocytogenesis and relapse, rather
than to differential sensitivity to antimalarial drugs, because
recent surveys showed no resistance of P. vivax to most an-
timalarial drugs except SP.
39
Current treatments for P. falciparum and P. vivax malaria
involve two different combinatory drug regimens to com-
bat multidrug resistance for P. falciparum (mefloquine and
artesunate) and relapse for P. vivax (CQ and primaquine).
13
Recent studies showed that P. vivax malaria relapse was
reduced to < 5% under the current antimalarial drug
policy compared with > 30% in late 1980s.
17,40–42
There-
fore, the recent P. vivax records may well represent the actual
vivax transmission in Thailand. Other studies have shown
that the recent drug policy was highly effective and helped
stabilize the multidrug resistance problems of P. falciparum
malaria in Thailand.
6,43–45
However, it must be empha-
sized that artesunate and other artemisinin combinations
are the last drugs in our line of defense in malaria chemo-
therapy. While significant drug resistance of P. vivax has not
been detected in Thailand, P. vivax malaria still prevails and
shows no signs of decreasing. In certain areas, such as Sa
Kaeo Province near the Thailand-Cambodian border, P.
vivax malaria has soared to an unprecedented level in the past
few years.
46
Since drug resistance in P. vivax has not been
detected in this area, the increase in P. vivax incidence rates
may be related to changes in the composition and abundance
of vectors with differential capabilities of transmitting the
two malaria parasites.
13,47
If this is true, the malaria control
program in Thailand will require an integrated approach
combining chemotherapy for the disease and control of the
vector.
ZHOU AND OTHERS260
Received February 26, 2004. Accepted for publication October 3,
2004.
Acknowledgments: We thank two anonymous reviewers for con-
structive suggestions.
Financial support: This work was supported by National Institute of
Health grant R01 AI-50243 and FIC D43 TW06571.
Authors’ addresses: Guofa Zhou and Guiyun Yan, Department of
Biological Sciences, 220 Hochstetter Hall, State University of New
York at Buffalo, Buffalo, NY 14260. Jeeraphat Sirichaisinthop, Vec-
tor Borne Disease Training Center, 6 Tambon TharnKasem,
Phrabuddhabat, Saraburi 18120, Thailand. Jetsumon Sattabongkot
and James Jones, Department of Entomology, Armed Forces Insti-
tute of Medical Sciences, United States Army Military Component,
Bangkok 10400, Thailand. Ottar N. Bjørnstad and Liwang Cui, De-
partment of Entomology, The Pennsylvania State University, Uni-
versity Park, PA 16802.
Reprint requests: Liwang Cui, Department of Entomology, The
Pennsylvania State University, University Park, PA 16802, Tele-
phone: 814-863-7663, Fax: 814-865-3048, E-mail: luc2@psu.edu or
Guiyun Yan, Department of Biological Sciences, 220 Hochstetter
Hall, State University of New York at Buffalo, Buffalo, NY 14260,
Telephone: 716-645-2363 extension 121, Fax: 716-645-2975, E-mail:
gyan@acsu.buffalo.edu.
REFERENCES
1. Somboon P, Lines J, Aramrattana A, Prajakwong S, Chitprarop
U, Khamboonruang C, 1995. Entomological evaluation of
community-wide use of lambdacyhalothrin-impregnated bed
nets against malaria in a border area of north-west Thailand.
Trans R Soc Trop Med Hyg 89: 248–254.
2. Ministry of Public Health, 2003. Malaria Control Program in
Thailand. Bangkok: Ministry of Public Health. http://eng.
moph.go.th/specifichealth/malaria/malaria.htm.
3. Somboon P, Aramrattana A, Lines J, Webber R, 1998. Entomo-
logical and epidemiological investigations of malaria transmis-
sion in relation to population movement in forest areas of
north-west Thailand. Southeast Asian J Trop Med Public
Health 29: 3–9.
4. Chareonviriyaphap T, Bangs MJ, Ratanatham S, 2000. Status of
malaria in Thailand. Southeast Asian J Trop Public Health 31:
225–237.
5. Ministry of Public Health, 2000. Thailand Health Profile, 1999–
2000. Bangkok: Ministry of Public Health. http://www.moph.
go.th/ops/thealth_44/
6. Wongsrichanalai C, Thimasarn K, Sirichaisinthop J, 2001. Anti-
malarial drug combination policy: a caveat. Lancet 355: 2245–
2247.
7. Singhanetra-Renard A, 1986. Population movement, socioeco-
nomic behavior and the transmission of malaria in northern
Thailand. Southeast Asian J Trop Med Public Health 17: 396–
405.
8. Fungladda W, Sornmani S, Klongkamunankarn K, Hungsapruek
T, 1987. Sociodemographic and behavioral factors associated
with malaria patients in Kanchanaburi, Thailand. J Trop Med
Hyg 90: 233–237.
9. Sornmant S, Butraporn P, Fungladda W, Okanurak K, Dissa-
pongsa S, 1983. Migration and disease problems: a study of
pattern of migration in an endemic area of malaria in Thailand.
Southeast Asian J Trop Med Public Health 14: 64–68.
10. Harinasuta T, Suntharasamai P, Viravan C, 1965. Chaloroquine
resistant falciparum malaria in Thailand. Lancet 2: 657–660.
11. Hall AP, Segal HE, Pearlman EJ, Phintuyothin P, Kosakal S,
1975. Amodiaquine resistant falciparum malaria in Thailand.
AmJTropMedHyg24:575–580.
12. Hurwitz ES, Johnson D, Campbell CC, 1981. Resistance of Plas-
modium falciparum malaria to sulfadoxine-pyrimethamine
(‘Fansidar’) in the refugee camp in Thailand. Lancet 1: 1068–
1070.
13. Sattabongkot J, Tsuboi T, Zollner GE, Sirichaisinthop J, Cui L,
2004. Plasmodium vivax transmission: chance for control?
Trends Parasitol 20: 192–198.
14. Maitland K, Willams TN, Newbold CI, 1997. Plasmodium vivax
and P. falciparum: Biological interactions and the possibility of
cross-species immunity. Parasitol Today 13: 227–231.
15. Maitland K, Williams TN, Bennett S, Newbold CI, Peto TEA,
Viji J, Timothy R, Clegg JB, Weatherall DJ, 1996. The inter-
action between Plasmodium falciparum and P. vivax in chil-
dren on Espirtu Santo island, Vanuatu. Trans R Soc Trop Med
Hyg 90: 614–620.
16. Bruce MC, Donnelly CA, Alpers MP, Galinski MR, Barnwell
JW, Walliker D, Day KP, 2000. Cross-species interac-
tions between malaria parasites in humans. Science 287: 845–
848.
17. Looareesuwan S, White NJ, Chittamas S, Bunnag D, Harinasuta
T, 1987. High rate of Plasmodium vivax relapse following
treatment of falciparum malaria in Thailand. Lancet 1: 1052–
1055.
18. Robinson TP, 2000. Spatial statistics and geographical informa-
tion system in epidemiology and public health. Adv Parasitol
47: 82–128.
19. Hay SI, Omumbo JA, Craig MH, Snow RW, 2000. Earth obser-
vation, geographic information systems and Plasmodium falci-
parum malaria in sub-Saharan Africa. Adv Parasitol 47: 174–
215.
20. National Statistical Office, 2000. Statistical Year Book Thailand.
Bangkok: National Statistical Office.
21. Asia-Pacific Economic Cooperation Committee (APEC), 2003.
APEC Economic Outlook. Singapore: APEC Secretariat.
22. Asia Collaborative Training Network for Malaria, 2003. Malaria
Profile, Thailand. Manila: The Philippines. http://www.
actmalaria.org/
23. Cliff AD, Ord JK, 1973. Spatial Autocorrelation. London: Pion.
24. Sokal RR, Oden NL, 1978. Spatial autocorrelation in biology 1.
Methodology. Biol J Linn Soc 10: 199–228.
25. Reynolds KM, Madden LV, 1988. Analysis of epidemics
using spatio-temporal autocorrelation. Phytopathology 78:
240–246.
26. Cliff AD, Ord JK, 1981. Spatial Process: Models and Applica-
tions. London: Pion.
27. Environmental Systems Research Institute, Inc., 1999. ArcView
GIS 3.2. Redlands, CA: Environmental Systems Research In-
stitute, Inc.
28. Lawson AB, 2001. Statistical Methods in Spatial Epidemiology.
New York: John Wiley.
29. Carpenter TE, 1988. Microcomputer programs for Markov and
modified Markov chain disease models. Prev Vet Med 5: 169–
179.
30. Neumann PJ, Araki SS, Arcelus A, Longo A, Papadopoulos G,
Kosik KS, Kuntz KM, Bhattacharjya A, 2001. Measuring Alz-
heimer’s disease progression with transition probabilities: es-
timates from CERAD. Neurology 57: 957–964.
31. Zhou G, Liebhold AM, 1995. Forecasting the spatial dynamics of
gypsy moth outbreaks using cellular transition models. Land-
scape Ecol 10: 177–189.
32. Zar JH, 1999. Biostatistical Analysis. Fourth edition. Englewood
Cliffs, NJ: Prentice Hall.
33. Silachamroom U, Krudsood S, Phophak N, Looareesuwan S,
2002. Management of malaria in Thailand. Korean J Parasitol
40: 1–7.
34. Wongsrichanalai C, Pickard AL, Wernsdorfer WH, Moshnick
SR, 2002. Epidemiology of drug-resistant malaria. Lancet In-
fect Dis 2: 209–218.
35. Cui L, Mascorro CN, Fan Q, Rzomp KA, Khuntirat B, Zhou G,
Chen H, Yan G, Sattabongkot J, 2003. Genetic diversity and
multiple infections of Plasmodium vivax malaria in western
Thailand. Am J Trop Med Hyg 68: 613–619.
36. Snounou G, Viriyakosol S, Jarra W, Thaithong S, Brown KN,
1993. Identification of the four human malaria parasite species
in field samples by the polymerase chain reaction and detec-
tion of a high prevalence of mixed infections. Mol Biochem
Parasitol 58: 283–292.
37. Panvisavas S, 2001. Poverty and malaria: a study in a Thai-
Myanmar border area. Southeast Asia J Trop Med Public
Health 32: 608–614.
38. Wongsrichanalai C, Sirichaisinthop J, Karwacki JJ, Congpuong
K, Miller RS, Pang L, Thimasarn K, 2002. Drug resistant ma-
SPATIO-TEMPORAL PATTERNS OF MALARIA INCIDENCE 261
laria on the Thai-Myanmar and Thai-Cambodian borders.
Southeast Asian J Trop Med Public Health 32: 41–49.
39. Pukrittayakamee S, Clemens R, Chantra A, Nontprasert A,
Luknam T, Looareesuwan S, White NJ, 2000. Therapeutic re-
sponses to antibacterial drugs in vivax malaria. Trans R Soc
Trop Med Hyg 95: 524–528.
40. Walsh DS, Looareesuwan S, Wilairatana P, Heppner DG Jr,
Tang DB, Brewer TG, Chokejindachai W, Viriyavejakul P,
Kyle DE, Milhous WK, Schuster BG, Horton J, Braitman DJ,
Brueckner RP, 1999. Randomized dose-ranging study of the
safety and efficacy of WR 238605 (tafenoquine) in the preven-
tion of relapse of Plasmodium vivax malaria in Thailand.
J Infect Dis 180: 1282–1287.
41. Wilairatana P, Silachamroon U, Krudsood S, Singhasivanon P,
Treeprasertsuk S, Bussaratid V, Phumratanaprapin W, Srivi-
lirit S, Looareesuwan S, 1999. Efficacy of primaquine regimens
for primaquine-resistant Plasmodium vivax malaria in Thai-
land. AmJTropMedHyg61:973–977.
42. Silachamroon U, Krudsoon S, Treeprasertsuk S, Wilairatana P,
Chalearmrult K, Mint HY, Maneekan P, White NJ, Gourdeuk
VR, Brittenham GM, Looareesuwan S, 2003. Clinical trial of
oral artesunate with or without high-dose primaquine for the
treatment of vivax malaria in Thailand. AmJTropMedHyg
69: 14–18.
43. Nosten F, van Vugt M, Price R, Luxemburger C, Rhway KL,
Brockman A, McGready R, ter Kuile F, Looareesuwan S,
2000. Effects of artesunate-mefloquine combination on incidence
of Plasmodium falciparum malaria and mefloquine resistance
in western Thailand: a prospective study. Lancet 356: 297–302.
44. van Vugt M, Looareesuwan S, Wilairatana P, McGready R, Vil-
legas L, Gathmann I, Mull R, Brockman A, White NJ, Nosten
F, 2000. Artemether-lumefantrine for the treatment of multi-
drug-resistant falciparum malaria. Trans R Soc Trop Med Hyg
94: 545–548.
45. Wilairatana P, Krudsood S, Chalermrut K, Pengruksa C, Sriv-
ilairit S, Silachamroon U, Treeprasertsuk S, Looareesuwan S,
2002. An open randomized clinical trial of Artecom vs artesu-
nate-mefloquine in the treatment of acute uncomplicated fal-
ciparum malaria in Thailand. Southeast Asian J Trop Med Pub-
lic Health 33: 519–524.
46. Limrat D, Rojruthai B, Apiwathnasorn C, Samung Y, Prommon-
gkol S, 2001. Anopheles barbirostris/campestris as a probable
vector of malaria in Aranyaprathet, Sa Kaeo Province. South-
east Asian J Trop Med Public Health 32: 739–744.
47. Somboon P, Suwonkerd W, Lines JD, 1994. Susceptibility of Thai
zoophilic anophelines and suspected malaria vectors to local
strains of human malaria parasites. Southeast Asian J Trop
Med Public Health 25: 766–770.
ZHOU AND OTHERS262
