Background : Malaria remains a dominant health issue among children in Ghana. We monitored the trend of long lasting insecticide net (LLIN) ownership and use and its impact on malaria and anaemia among children under five over the past decade in an area of intense, prolonged and seasonal malaria transmission. Methods : A total of 1717, 2155 and 1915 children were surveyed in June and 1717, 2155 and 1697 in November in 30 communities of the Hohoe Municipality in 2006, 2010 and 2015 respectively. The primary outcomes monitored were the trend of LLIN ownership, use and prevalence of malaria and anaemia through cross-sectional surveys. Findings : Results showed a significantly positive trend (p < 0.001) in ownership of LLIN at 20.8%, 64.8% and 83.2% for 2006, 2010 and 2016 respectively. The sharp increase in ownership between 2006 and 2010 accounted for a highly significant departure from trend (p < 0.001). Similarly, the prevalence of LLIN use was 15.1%, 42.2% and 68.3% for 2006, 2010 and 2016 respectively with a significant positive trend (p < 0.001). The prevalence of malaria parasitaemia in June showed no significant drop over the three years studied. Anaemia prevalence was 7.9%, 5.3% and 11.1% for 2006, 2010 and 2015 respectively—Anaemia significantly increased (p < 0.001) between 2010 and 2015 accounting for a significant departure from the initially decreasing trend (p < 0.001). Malaria parasitaemia over the November surveys was 40.4%, 33.2% and 26.6% for 2006, 2010 and 2016 respectively, depicting a significant decline over the years of the study (p < 0.001). A significant decrease (p = 0.006) in anaemia was observed at 12.0%, 4.3% and 8.9% for 2006, 2010 and 2015 respectively. A significant departure from the decreasing trend (p < 0.001) was noted with increasing anaemia between 2010 and 2016. Interpretation : Ownership of LLIN and its use together with other indicators of malaria prevalence generally improved five years after the implementation of a malaria control programme. Ten years after programme implementation, consistent improvement was only in LLIN ownership and use. Also malaria prevalence indicators improved in the post as compared to the pre-rainy season in the study communities.
Malaria is a focal problem of sub-Saharan Africa and for that matter Ghana. The condition is endemic in Ghana with Plasmodium falciparum (pf) as the predominant species causing the disease [
Though indicators point to a drop in its occurrence, malaria remains a dominant health issue in Ghana. Some 2200 deaths out of about 3.5 million cases of malaria reported in Ghana in 2015 were significantly lower than figures from 2013 [
The Millennium Development Goals (MDGs) and the Abuja Declaration set targets of achieving 60% coverage of malaria interventions in pregnant women and children below 5 years of age by 2010 [
Methods employed by malaria control programmes globally include the use of long lasting insecticide nets (LLINs), and early detection of malaria and appropriate and prompt management with artemisinin-based combination therapy (ACT) [
Important roll back malaria (RBM) indicators that help in monitoring changes in malaria incidence and prevalence are the proportion of households that own one or more LLINs and the proportion of under-five children who sleep inside a LLIN [
In line with attaining MDG targets of reducing malaria illness and deaths, the Ghana NMCP introduced LLINs, IPTs and ACTs into the Hohoe Municipality. Ahead of implementing the intervention, baseline surveys were carried out in June and November 2006 [
The design of the three surveys was cross-sectional in nature. The surveys were carried out at the beginning of the rainy season (pre-rainy season) in June and the end (post-rainy season) in November) of the high transmission period for malaria in 2006, 2010 and 2015 respectively.
The study was undertaken in the Hohoe Municipality of Ghana. The municipality is one of the twenty-five administrative districts of the Volta Region with a population of 167,000 people. It has two main seasons―the wet and the dry. The major wet season is from April to July and the minor one from September to November each year. The rest of the year is relatively dry. The climate is tropical with temperatures varying between 22˚C and 37˚C. The average annual rainfall in the municipality is 1592 mm with approximately 1296 mm of rain falling between April and October [
The study population was all children in the municipality aged 6 to 59 months. To be eligible for the study, the child should be aged 6 - 59 months and the parents/guardians consented to participate.
For the initial survey in 2006, thirty communities were selected from a sample frame of all villages and communities in the municipality. The number of children living in each community was determined from the 2000 census. Thirty communities with the required number of children for the study (2125) were randomly selected by probability proportional to the number of children living in each community. This sampling approach was used to ensure a fair representation of communities in the municipality in the study. The sample size was estimated on the basis of the following: 95% confidence level (Z) and power of 80%, the prevalence (P) of malaria in children aged less than 5 years in June 2006 (end of dry season) was 8.6%. The least acceptable prevalence of malaria was 5.0%. Using OpenEpi software version 3 [
Data collection for each study was in three phases:
Phase 1: Parents/guardians of the sampled children were interviewed using a close-ended questionnaire to gather information among others on LLIN ownership and use.
Phase 2: Temperature and weight measurements were taken of the sampled children. Temperature was measured using electronic thermometers (MODE: ZC, SURGILAC Digital Thermometer, UK). Weight was measured using SECA weighing scales (Hamburg, Germany).
Phase 3: Finger-prick blood samples were collected from the sampled children to assess malaria parasitaemia (mps) and Hb levels. Two technicians examined thick blood films for mps. A sample was considered negative only after 200 high power fields had been read. Parasite counts were converted to parasites per micro-liter (μl), assuming a white blood cell count of 8000 leukocytes per μl of blood. If there was a discrepancy in the findings of a slide between the two technicians (positive or negative or a 50% or more difference in parasite density) a third, more senior microscopist provided a conclusive result. A senior microscopist from the School of Public Health of the University of Health and Allied Sciences examined all the positive blood films and a 20% random sample of negative blood slides for quality control. Haemoglobin was measured using URIT-12 Hemoglobin Meter (URIT Medical Electronic, UK).
Data collected from the study subjects were recorded on specified forms and checked by field supervisors and a data manager for consistency and accuracy. The data were double entered into a database using the EPI Data software version 3.1. The accuracy of data input was checked and validated using customized validation programmes. The cleaned data were exported into the STATA statistical software version 12 for analysis.
The main variables of interest were malaria prevalence measured by any malaria parasitaemia (asexual parasite count 40/μl and above), high density parasitaemia (asexual parasite count ≥7000/μl, gametocytemia (sexual parasites), fever (axillary temperature ≥ 37.5˚C) and anaemia (Hb < 8.0 g/dl) [
Analysis of the outcomes was carried out on an intention-to-treat basis, which implies that children screened were included in the analysis. Pre-rainy and post-rainy season measures of malaria parasitaemia, gametocytaemia, fever and anaemia in 2010 and 2015 were compared with those of the same periods of 2006. For each outcome a regression of its proportion was run using the Stata command ptrend. It reports the slope, standard error and z-score of the regression line, a chi square test for trend and a chi square test for departure from trend together with their corresponding p-values. All analyses were done with STATA software version 12.0. A negative slope was referred to as a negative trend, which implies a reduction in prevalence compared to the previous years. A positive slope was referred to as a positive trend, which implies an increase in prevalence over the previous years.
Ethical approval for the study was obtained from the Ethical Review Committee (ERC) of the Ghana Health Service/Ministry of Health (GHS/MOH). The identification number for the clearance was ETHICS APPROVAL-ID NO: GHS-ERC: 14/05/15. Before the commencement of the study, permission was sought from the Municipal Health Management Team (MHMT), the Municipal Administration and the chiefs and elders of the selected communities. Parents/guardians of the participating children signed a written informed consent form before the start of the study.
The sample population of the study in 2015 was one thousand nine hundred and fifteen (1915) and one thousand six hundred and ninety-seven (1697) children less than five years in the pre-rainy and post-rainy seasons respectively. The population for June and November 2006 was 1717 and 1435 respectively, and that for 2010, 2155 and 1778 children respectively. Overall, sex, mean age and mean weight distribution were similar in all groups (
Variable | Pre-rainy season | ||
---|---|---|---|
Age groups (in months) | 2006 (N = 1717) [n (%)] | 2010 (N = 2155) [n (%)] | 2015 (N = 1915) [n (%)] |
6 - 11 | 321 (18.7) | 389 (18.0) | 258 (13.5) |
13 - 23 | 435 (25.3) | 510 (23.7) | 428 (22.3) |
24 - 35 | 399 (23.2) | 454 (21.1) | 480 (25.1) |
36 - 47 | 319 (18.6) | 436 (20.2) | 377 (19.7) |
48 - 59 | 243 (14.2) | 366 (17.0) | 372 (19.4) |
Mean weight (kg) [SD] | 12.8 (2.7) | 11.3 (3.3) | 11.8 (3.2) |
Mean age (in months) [SD] | 27.3 (15.4) | 27.8 (15.5) | 30.6 (15.6) |
Variable | 2006 (N = 1717) [n (%)] | 2010 (N = 2155) [n (%)] | 2015 (N = 1915) [n (%)] | X 1 2 , P (Trend) | X 1 2 , P (Depart.) |
---|---|---|---|---|---|
Own LLIN | 357 (20.8) | 1396 (64.8) | 1593 (83.2) | 1422, < 0.001a | 90.6, <0.001 |
Use LLIN | 260 (15.1) | 910 (42.2) | 1307 (68.3) | 2.1, < 0.001b | 0.155, 0.693 |
a: Slope = 0.309, std. error = 0.008, z = 37.72; b: Slope = 0.265, std. error = 0.008, z = 32.30.
Variable | 2006 (N = 1717) [n (%)] | 2010 (N = 2155) [n (%)] | 2015 (N = 1915) [n (%)] | X 1 2 , P (Trend) | X 1 2 , P (Depart.) |
---|---|---|---|---|---|
Fever (temp. ≥ 37.5) | 42 (2.4) | 25 (1.2) | 50 (2.6) | 0.2, 0.637a | 12.8, <0.001 |
Any malaria parasitaemia | 156 (9.1) | 156 (7.2) | 148 (7.7) | 2.1, 0.145b | 2.5, 0.113 |
High density malaria parasitaemia | 50 (2.9) | 49 (2.3) | 43 (2.2) | 1.6, 0.203c | 0.5, 0.469 |
Gametocytaemia | 14 (0.8) | 3 (0.1) | 7 (0.4) | 4.1, 0.043d | 6.7, 0.010 |
Anaemia (Hb < 8.0 g/dl) | 135 (7.9) | 114 (5.3) | 212 (11.1) | 14.1, <0.001e | 32.1, <0.001 |
a: Slope = 0.001, std. error = 0.002, z = 0.472; b: Slope = −0.002, std. error = 0.001, z = 2.021; c: Slope = −0.003, std. error = 0.003, z = 1.273; d: Slope = −0.002, std. error = 0.001, z = 2.021; e: Slope = 0.017, std. error = 0.005, z = 3.756.
In
The prevalence of “any parasitaemia” was 9.1%, 7.2% and 7.7% respectively for 2006, 2010 and 2016. The test of trend between 2006 and 2015 showed that even though there was a negative trend, it was not significant (p = 0.145).
The prevalence of “high density parasitaemia” was 2.9%, 2.3% and 2.2% respectively for 2006, 2010 and 2015. The test of trend between 2006 and 2015 showed that even though there was a negative trend, it was not significant (p = 0.203).
The prevalence of gametocytemia was 0.8%, 0.1% and 0.4% respectively for 2006, 2010 and 2015. The test of trend between 2006 and 2015 showed that there was a negative trend and it was significant (p = 0.043). The drop in prevalence between 2006 and 2010 and the rise in prevalence between 2010 and 2015 was a significant departure from the trend (p = 0.010).
The prevalence of anaemia was 7.9%, 5.3% and 11.1% respectively for 2006, 2010 and 2015. The test of trend between 2006 and 2015 showed a significant rise in cases of anaemia (p < 0.001). Even though there was a significant fall between 2006 and 2010, the rise between 2010 and 2015 was a significant departure from the trend (p < 0.001).
In
Variable | 2006 (N = 1717) [n (%)] | 2010 (N = 2155) [n (%)] | 2015 (N = 1915) [n (%)] | X 1 2 , P (Trend) | X 1 2 , P (Depart.) |
---|---|---|---|---|---|
Fever (temp. ≥ 37.5) | 71 (4.9) | 77 (4.3) | 44 (2.6) | 11.8, 0.001 | 0.9, 0.331 |
Any malaria parasitaemia | 580 (40.4) | 590 (33.2) | 451 (26.6) | 67.3, <0.001 | 0.05, 0.822 |
High density malaria parasitaemia | 151 (10.5) | 56 (3.1) | 71 (4.2) | 54.0, <0.001 | 37.4, <0.001 |
Gametocytaemia | 12 (0.8) | 3 (0.2) | 5 (0.3) | 5.1, 0.023 | 4.4, 0.036 |
Anaemia (Hb <8.0 g/dl) | 172 (12.0) | 76 (4.3) | 151 (8.9) | 7.7, <0.006 | 57.6, <0.001 |
a: Slope = −0.012, std. error = 0.003, z = 3.441; b: Slope = −0.069, std. error = 0.008, z = 8.207; c: Slope = −0.030, std. error = 0.001, z = 2.268; d: Slope = −0.003, std. error = 0.001, z = 2.257; e: Slope = −0.014, std. error = 0.005, z = 2.772.
2015 showed a significant reduction of cases of fever over the years studied (p = 0.001).
The prevalence of “any parasitaemia” was 40.4%, 33.2% and 26.6% respectively for 2006, 2010 and 2016. The test of trend between 2006 and 2015 showed a significant reduction of “any parasitaemia” over the years of the study (p < 0.001).
The prevalence of “high density parasitaemia” was 10.5%, 3.1% and 4.2% respectively for 2006, 2010 and 2015. The test of trend between 2006 and 2015 showed a significant decrease in “high density parasitaemia” over the years studied (p < 0.001). However, there was a fall between 2006 and 2010 and then a rise between 2010 and 2016 but not up to the level of 2006. The fall in prevalence between 2006 and 2010 and the rise between 2010 and 2015 formed a significant departure from the linear trend (p < 0.001) (
The prevalence of “gametocemia” was 0.8%, 0.2% and 0.3% respectively for 2006, 2010 and 2016. The test of trend between 2006 and 2015 showed a significant reduction over the study period (p = 0.023). The fall in prevalence between 2006 and 2010 and the rise between 2010 and 2015 formed a significant departure from the linear trend (p = 0.036) (
The prevalence of anaemia was 12.0%, 4.3% and 8.9% respectively for 2006, 2010 and 2015. The test of trend between 2006 and 2015 showed a significant decrease in cases of anaemia (p = 0.006). The fall in prevalence between 2006 and 2010 and the rise between 2010 and 2016 formed a significant departure from the linear trend (p < 0.001) (
The study determined changes in indicators related to malaria prevalence among children below five years of age in the Hohoe Municipality of Ghana after interventions were introduced in 2006. This study has the unique quality of two five-year follow-ups post intervention. This has provided a longer-term view of the intervention effect in its immediate and long-term aftermath. There was a general increase in LLIN ownership and use. LLIN use consistently lagged behind LLIN ownership. Malaria parasitaemia, fever and anaemia levels fell within the first five years of the malaria control interventions, but were not sustained five years on.
This study shows an increase in LLIN ownership and use in the Hohoe municipality between 2006 and 2015 in the pre-rainy season. The findings are similar to surveys carried out in 2003, 2008 and 2014 that showed a consistent increase in ownership and use of LLINs in Ghana as a whole and in the Volta Region in particular (Ghana Demographic and Health Survey [GDHS], 2008 & 2014). Just as in the current study population, LLIN use has been consistently lower than its ownership across populations in some earlier studies [
Significant reductions in pf and anaemia were observed in children below 5 years after introduction of LLIN in Zanzibar (p < 0.001) [
The increasing LLIN ownership and use observed in this study did correspond to a continuous reduction in malaria prevalence during the high transmission but not the low transmission season. Studies in Kenya also made similar observations [
The effectiveness of LLIN reduces over time. However, the net does not need to be sprayed periodically with insecticide over time to maintain its effectiveness. After a period of time, the net needs to be replaced. This fact could be the reason why despite an increase in LLIN ownership and usage in the current study, there was still an increase in gametocytaemia and anaemia. However, information on maintenance of LLIN was not collected from the net owners during the current study.
Limitations of the StudyThis was a cross-sectional survey at one point of the end of the high transmission season and not a cohort study that can determine malaria incidence throughout the wet and dry seasons. Information on LLIN ownership and usage was obtained from parents of the children without observation to confirm the availability and use. Since this was not a cohort study, our interest was not to study a particular sample of children over the period of the study. However, there might be some overlap between the selected samples over the period of the study as measures were not put in place to ensure that different sets of children were evaluated in the three different periods of the study.
Long Lasting Insecticide Treated Net ownership and use together with other indicators of malaria prevalence generally improved five years after the implementation of a malaria control programme in the Hohoe Municipality. Ten years after program implementation consistent improvement was only in LLIN ownership and use. Also malaria prevalence indicators improved better in the post as compared to the pre-rainy season.
Considering the fact that LLIN ownership and use showed consistent improvement, there is the need to further investigate factors that contributed to other malaria prevalence indicators. In the meantime other measures such as seasonal malaria chemoprevention (SMC) and indoor insecticide paints among others should be employed by the Municipal Health Directorate in consultation with other stakeholders to improve malaria control indicators.
We are grateful to Dr. Felix Doe, the Hohoe Municipal Director of Health Services and Municipal Health Management Team for their assistance during the survey. We are also grateful to Mr. King Kpo, who monitored the quality of Hb measurement, prepared slides and read them. We are thankful to all the parents/guardians for consenting to them and their children to be part of the study.
All authors report no conflict of interest in this study.
MK worked on the 2006 and 2015 projects and 2010 post-rainy season survey, supported EA on the 2010 pre-rainy season study who worked on this project for his MPH. MK, EA conceived the study. MK, YE, MA, WT, ET and JG were responsible for the initial draft of the manuscript. All authors reviewed and approved the final version of the manuscript.
Kweku, M., Appiah, E.K., Enuameh, Y., Adjuik, M., Takramah, W., Tarkang, E. and Gyapong, J. (2017) The Impact of Malaria Control Interventions on Malaria and Anaemia in Children under Five after Ten Years of Implementation in the Hohoe Municipality of Ghana. Advances in Infectious Diseases, 7, 93-105. https://doi.org/10.4236/aid.2017.73010