Background
There have been previous efforts to eradicate or, if this was not possible, to eliminate malaria, but these ambitious goals have not been achieved in large parts of Africa [
1‐
5] despite the development of effective tools, such as insecticide-impregnated bed nets (ITNs) and case management on demand; and malaria vaccines are being developed to complement the existing tools [
6‐
9]. The performance of these tools can be monitored by careful surveillance of the malaria burden in a defined community before and after their deployment [
10‐
12]. However, malaria transmission and the associated disease burden can fluctuate in the absence of any intervention due to climatic variations and societal behavioural changes. These changes can be dramatic in areas of unstable and seasonal malaria transmission, but are much less pronounced in areas of high and stable transmission [
13,
14]. In the past few years, reports have documented that the malaria burden has diminished in several areas of Africa [
15‐
18], but the reasons for these changes are still uncertain. This fortunate development has been ascribed to a combination of interventions, such as use of ITNs, artemisinin combination therapy (ACT) and indoor residual spraying (IRS) [
16]. This paper documents a marked decline in the malaria burden in two Tanzanian villages in an area with stable malaria transmission during the period 2003 to 2008. The decrease was most marked in a village where transmission was high at the beginning of the study and took place before the introduction of ACT.
Methods
Study area
The study was conducted in Korogwe District, north-eastern Tanzania. Korogwe District is about 100 kilometers inland from the coastal town, Tanga. It is a tropical area with two rainy seasons: April to June and October to December. January and February are normally dry, but recently there have been climatic changes with merging of these two rainy seasons. Entomological surveys in the study area have shown that
Anopheles gambiae is the most prevalent vector in the lowlands, while
Anopheles funestus predominates in the highlands. In previous studies, entomological inoculation rates (EIR) have been reported to be more than 100 in the lowlands, but less than 30 infective bites per person per year at intermediate altitudes. In the highlands, mosquito densities are too low to allow reliable EIR measurements, but an EIR of 0.03 infective bites per person per year has been extrapolated [
19]. There is no recent entomology data for the area, the most recent were collected in the late 1990 s.
Plasmodium falciparum is the dominant malaria parasite with
Plasmodium malariae being encountered occasionally. Villagers perceive malaria as a major health problem among both children and adults leading to disease and deaths.
The study was conducted in two villages, Mkokola and Kwamasimba, which had 2,000 and 1,800 inhabitants, respectively in 2003. Mkokola village is situated at an altitude of about 300 meters above sea level (lowland village) and Kwamasimba at an altitude of approximately 700 meters (highland village). The two villages are approximately 15 kilometres apart [
20‐
22].
Longitudinal detection of febrile episodes
In January 2003, management of uncomplicated malaria cases by trained village health workers was introduced into the study villages. The village health workers, locally called Community Owned Resource Persons (CORPS) carried out passive case detection of subjects with fever. There were two CORPS in each village, one of whom was always present in the village, so that any village resident who had a febrile illness, had an opportunity to obtain treatment at any time. The CORPS were provided with thermometer, first-line anti-malarial drugs (sulphadoxine-pyrimethamine [SP] up to 2006 and artemether-lumefantrine [ALu] from 2007 to date), paracetamol, slides, blood lancets, gloves, treatment charts, febrile case detection forms and storage boxes. CORPS were instructed how to collect blood smears under aseptic technique and to preserve these until collected by a team of supervisors composed of a clinical officer and a laboratory technician. CORPS were also instructed on how to treat patients with the first-line anti-malarial drug and an antipyretic, and how to refer patients who had severe symptoms or who did not respond adequately to the first-line anti-malarial treatment to Korogwe District Hospital. Villagers were informed that they could call on the CORPS at any time if they had symptoms of malaria. From suspected malaria cases, the CORPS would then collect basic clinical data, obtain a blood smear for malaria microscopy and institute presumptive malaria treatment. Patients with severe symptoms or symptoms suggestive of other diseases than malaria were referred to the local dispensary or the District Hospital.
Cross-sectional surveys
Malaria cross-sectional surveys were conducted in the two study villages from 2003 to 2008 (Table
1). The surveys were conducted in May during the main rainy season in the years 2003-2008, and just before the short rains in November in the years 2003-2007. During each cross-sectional survey, demographic data were collected together with a history of migration and recent travel. The reported use of a bed net was also recorded. A history of recent illness was obtained, emphasizing symptoms suggestive of malaria. Physical examination for signs related to malaria, such as temperature, pulse, spleen size, pallor or a raised respiratory rate was conducted. Axillary temperature was measured using digital thermometers. For any individual diagnosed with a mild disease, appropriate drugs were administered.
Table 1
Number of participants surveyed per village per year
Lowland village
|
<1
| 30 | 15 | 20 | 30 | 23 | 5 |
|
1
| 35 | 48 | 55 | 34 | 35 | 21 |
|
2
| 31 | 63 | 61 | 52 | 14 | 20 |
|
3
| 29 | 45 | 79 | 52 | 20 | 7 |
|
4
| 26 | 47 | 63 | 70 | 23 | 20 |
|
5-9
| 76 | 157 | 191 | 212 | 130 | 59 |
|
10-14
| 47 | 65 | 153 | 56 | 165 | 49 |
|
15-19
| 36 | 52 | 55 | 18 | 58 | 22 |
|
Total
|
310
|
492
|
677
|
524
|
468
|
203
|
Highland village
|
<1
| 25 | 23 | 42 | 33 | 34 | 5 |
|
1
| 25 | 51 | 80 | 50 | 44 | 14 |
|
2
| 20 | 53 | 73 | 72 | 40 | 17 |
|
3
| 29 | 32 | 77 | 47 | 60 | 15 |
|
4
| 28 | 56 | 58 | 67 | 48 | 20 |
|
5-9
| 81 | 160 | 123 | 220 | 207 | 55 |
|
10-14
| 44 | 63 | 75 | 90 | 141 | 52 |
|
15-19
| 31 | 54 | 46 | 28 | 40 | 34 |
|
Total
|
283
|
492
|
574
|
607
|
614
|
212
|
Five millilitres of venous blood were collected from individuals aged three years or more. For children below three years, 300-400 μl of capillary blood was obtained by finger prick into an Eppendorf tube containing EDTA. The haemoglobin (Hb) concentration of each participant was measured using a HemoCue® photometer (Ångelholm, Sweden). Whole blood was used to prepare thick and thin blood smears for malaria microscopy. These were stained with 10% Giemsa stain for 15-20 minutes after fixing thin smears with methanol. Asexual and sexual parasites were counted against 200 and 500 white blood cells, respectively. The differentiation of malaria parasite species was confirmed by microscopy of thin smears. A blood smear was declared negative only after examination of 200 high power fields. The density of asexual parasites was calculated assuming 8,000 leucocytes per μl of blood and expressed as parasites per μl. Slides were red retrospectively and during surveys malaria treatment was only given to patients complaining of malaria symptoms.
Data management and analysis
All data were double-entered into a Microsoft Access database, cleaned, validated and transferred into Stata version 9 (Stata Corporation, Texas, USA) and R 2.7.0 for statistical analysis. Fever was defined as an observed axillary temperature of ≥37.5°C and/or a history of fever within the previous 24 hours. A febrile malaria episode was defined as an episode of fever with a positive blood smear for asexual Plasmodium falciparum parasitaemia of any density. The summary measures consisted of percentage for categorical variables and means for continuous variables. Incidence was calculated as documented number of febrile malaria episodes to the population size in each age group and/or village per year.
In modelling the risk of Plasmodium falciparum, parasiteamia across the years, number of individuals with parasitaemia during cross-sectional surveys were modelled as a binomial random variable using logistic regression. The model was fitted for each village separately and was adjusted for the effect of age and use of bednets. To compare the risk of Plasmodium falciparum infection between the two villages, a similar model was fitted for the two villages. In addition, linear regression models were used to determine the changes in mean haemoglobin (Hb) concentration across years and effect of Plasmodium falciparum parasiteamia and season on Hb levels while controlling for age and reported use of insecticide-treated or non-treated bed nets. In this model the effect of year was fitted as a categorical variable. A statistical significance level was considered at P-value < 0.05.
Ethical considerations
Ethical clearance was granted by the Medical Research Co-ordinating Committee of the National Institute for Medical Research, Tanzania. Prior to the study, meetings were held with local authorities and with the villagers in each study village, during which the aims of the study were explained. Informed consent documents for the study were prepared in English and translated into Kiswahili before administration to both village leaders and participants. Written informed consent to participate was obtained from each study individual or from his or her parent or guardian. Villagers were free not to participate in the study without giving any reasons, or being disqualified from any medical services that were provided to all villagers throughout the study period. Feedback to the village has been performed continuously during the study.
Discussion
This community based study assessed the malaria burden in two villages in Korogwe district, north-eastern Tanzania by collecting data in repeated cross-sectional surveys and by employing CORPS to monitor malaria morbidity over six consecutive years. In this study, lowland residents had almost a twelve-fold increase in the risk of malaria parasitaemia compared to their counterparts in the highland village. At the start of the study, the lowland village could be categorized as high transmission whereas transmission in the highland village could be categorized as moderate. During the six years of the study, there was a marked and progressive decline in the malaria burden and this was reflected in the point prevalence of parasitaemia, the incidence of clinical episodes of malaria and in the mean haemoglobin concentration.
It is well established that immunity to malaria is a function of age, and that in areas of high transmission the malaria burden mainly is carried mainly by children and infants, while in areas of low transmission, all age groups are affected [
23]. As malaria transmission declined in the study area, a change in the age distribution of the malaria burden was detected. In 2003, young children resident in the high transmission village had a markedly higher incidence of malaria fevers than older children and young adults, and many of these children were anaemic. Indeed, the most striking finding of this study was the pronounced increase in the haemoglobin concentrations among 1-3 year old children in the lowland village. This represents a remarkable health gain and underscores findings of previous studies showing the high impact malaria has on haemoglobin levels in young children in areas of high transmission [
15,
24,
25]
During the study, the malaria burden also decreased among inhabitants of highland village and among the older age groups in lowland village, but because the decline in burden was more pronounced among the young children in lowland village, the net effect was that, at the end of the study the malaria burden appeared to be more equally distributed among the different age groups. During the study, many cases of severe malaria were not documented. This is probably due to the fact that a relatively low number of individuals were surveyed, and that these individuals had access to prompt and effective treatment. Thus, how the decline in transmission has affected the incidence of severe malaria in its different manifestations could not be estimated. This, is an important question as cerebral malaria has been reported to affect slightly older children than those affected by severe anaemia, and a declining incidence might lead to an increase in this severe form of malaria [
14,
26].
Other studies mainly based on data collected at health facilities have documented that malaria transmission seems to be falling in other parts of East Africa and in some of these studies the decrease has been attributed to specific interventions [
14‐
18]. In this community based study, a dramatic decrease in malaria indices in the absence of specific interventions other than giving villagers access to treatment through trained village helpers is documented. During the first four years of the study period, the first-line anti-malarial was SP, but from January 2007 an ACT was used. Thus, the most dramatic decrease in malaria burden took place when SP, a drug to which there is widespread resistance in the area, was in use [
27]. Although the presumptive treatment of febrile cases could have played a role in reducing transmission[
28], this could not explain the findings. In 2005, World Vision deployed many nets in Korogwe, including the study villages. Although the coverage was not complete, there has been a substantial increase in use of nets in the study villages, approaching 60%. It has been argued that communal effect of protection is achieved when coverage reaches 64%, which is almost the case in these villages [
26], but again whether nets at the coverage reached could have had such pronounced effects on transmission as indicated by the decrease in point prevalence from 2003 to 2008 is still debatable. As in Tanzania in general, there has been a steady socio-economic development in the area, but there have not been major changes or investments in infrastructure in the two study villages that can explain the decrease in malaria morbidity.
There is no simple explanation for the decline in malaria in the study villages, and it is rather an unsatisfactory situation of having to assign this to a combination of different factors such as climatic changes, better health provision, increased coverage of bed nets and socio economical development. It is very important to establish the relative contribution of these and other factors to the decline in malaria, as this will shed light on the risk that malaria might strike back. Entomological studies in the study villages were not done during the study period and whether there have been changes in mosquitoes distribution across years is not established.
In conclusion, there has been a dramatic progressive decline in point prevalence of malaria in the study villages during the study period, and in parallel the malaria disease burden has fallen, especially among those most affected previously, the infants and young children living in the village where malaria transmission was high previously.
Acknowledgements
The authors are grateful to the dedicated study teams which conducted cross-sectional surveys through data collection and processing. They also acknowledge discussion made and contributed by Michael Alifrangis, Pamela Magistrado, Samuel Gesase, Daniel Minja, Chris Drakeley and Deus Ishengoma. Technical and logistics supports of Ezekiel Malecela, Juma Akida, Dominick Shauri, Thomson Mwampamba, Rashid Madebe, Tilaus Gustav, Johari Sadi, Sia Minja, Eva Nyenza, Silas Msangi, Aneth Goso, Stella Mkandawile, Julius Mhache and Allan Mrango are highly appreciated. Authors remain indebted to the study participants for their time and endurance throughout the process. Brian Greenwood is acknowledged for reviewing the final manuscript. DANIDA and Gates Malaria Partnership are acknowledged for financial support.
Competing interests
The authors declare that they have no competing interests.
Authors' contributions
All authors were involved in the design of the study. LV, BM, AK, ML, TT and JL conceived the study. BM, LV and JL and were responsible for the implementation of the study in the field. BM coded the data and supervised data entry. BM and JL analysed data. JL wrote the first draft of the manuscript. All authors were involved in revising the manuscript, and read and approved the final version.