New PhD by Alemayehu Hailu: Economic evaluation of malaria prevention in Ethiopia

Thesis_UiB_Alemayehu_Hailu

Hailu A, D. Economic evaluation of malaria prevention in Ethiopia: Economic burden, equity, and cost-effectiveness analysis of malaria prevention in south-central Ethiopia [PhD]. Bergen: University of Bergen; 2018.

Summary
Background: Despite remarkable efforts in the global fight against malaria and achievements in the reduction of morbidity and mortality in the last decade, the disease remains to be a huge challenge to the health systems of malaria-endemic low-income countries in Africa and in all corners of the globe. Beyond the wide range of consensus on the disease burden and prioritization of malaria, the available evidence on the economic burden of malaria in Ethiopia is scanty. No clear evidence yet exists about the additional resources required for a combined implementation of long-lasting insecticidal nets (LLINs) and indoor residual spraying (IRS) with respect to the added protection afforded. In addition, little is known about the prevailing status of LLINs and IRS across different levels of wealth strata.

Objectives: The aims of this thesis were to estimate the economic burden of malaria; to evaluate the cost-effectiveness of a combined implementation malaria prevention interventions (LLINs and IRS); and analyse the distributional (equity) implications of the interventions in the Adami Tullu district, south-central Ethiopia.

Methods: Studies included in this thesis were conducted from 2014 – 2016 in the Adami Tullu district of Oromia Region, in south-central Ethiopia as part of and partly in conjunction with the MalTrials project. We conducted a cost of illness using 190 malaria patients in the first study. In Paper II, we carried out a combination of trial-based and literature-based cost-effectiveness analysis using Markov modelling. In the third study, using a cross-sectional data from 6069 households we did an inequality analysis of ownership of LLINs and IRS status. The principal component analysis technique was used for ranking households based on socioeconomic position. We measured the inequality in LLINs and IRS using concentration indices and concentration curve (Paper I&III).

Results: The median cost of malaria per episode to the household was USD 5.06 (IQR: 2.98 – 8.10) and the direct cost was significantly higher among the poor. The trial-based analysis had shown that the routine practice dominates both the combined intervention and singleton intervention while the literature-based analysis had indicated that combined intervention had an Incremental Cost-Effectiveness Ratio of USD 1403 per DALY averted. Immediately before we started the trial, the LLIN ownership was 11.6% and IRS coverage was 72.5%. We found a concentration index of 0.0627 for LLINs and – 0.0383 for the IRS. Inequality in LLIN ownership was mainly associated with a variability in a housing situation, the size of the household, and access to mass media and telecommunication service.

Conclusions: The economic burden of malaria to the rural households in Ethiopia is huge—mainly to the poor. Based on the trial-based cost-effectiveness analysis, we conclude that the combination of LLINs and IRS is not likely to be a cost-effective option compared with singleton intervention. However, based on the literature-based analysis, the combined intervention had potential to be a cost-effective alternative at 3 times GDP per capita per DALY averted. Furthermore, the ownership of LLIN was very low and significantly pro-rich, while IRS status was equitable across socioeconomic strata.

New and important publication from the Arba Minch group

Mulchandani R, Massebo F, Bocho F, Jeffries CL, Walker T, Messenger LA. A community-level investigation following a yellow fever virus outbreak in South Omo Zone, South-West Ethiopia. PeerJ. 2019;7. doi: 10.7717/peerj.6466.

Background

Despite the availability of a highly effective vaccine, yellow fever virus (YFV) remains an important public health problem across Africa and South America due to its high case-fatality rate. This study investigated the historical epidemiology and contemporary entomological and social determinants of a YFV outbreak in South Omo Zone (SOZ), Ethiopia.

Methods

A YFV outbreak occurred in SOZ, Ethiopia in 2012–2014. Historical epidemiological data were retrieved from the SOZ Health Department and analyzed. Entomological sampling was undertaken in 2017, including mosquito species identification and molecular screening for arboviruses to understand mosquito habitat distribution, and finally current knowledge, attitudes and preventative practices within the affected communities were assessed.

Results

From October 2012 to March 2014, 165 suspected cases and 62 deaths were reported, principally in rural areas of South Ari region (83.6%). The majority of patients were 15–44 years old (75.8%) and most case deaths were males (76%). Between June and August 2017, 688 containers were sampled across 180 households to identify key breeding sites for Aedesmosquitoes. Ensete ventricosum (“false banana”) and clay pots outside the home were the most productive natural and artificial breeding sites, respectively. Entomological risk indices classified most sites as “high risk” for future outbreaks under current World Health Organization criteria. Adult mosquitoes in houses were identified as members of the Aedes simpsonicomplex but no YFV or other arboviruses were detected by PCR. The majority of community members had heard of YFV, however few activities were undertaken to actively reduce mosquito breeding sites.

Discussion

Study results highlight the potential role vector control could play in mitigating local disease transmission and emphasize the urgent need to strengthen disease surveillance systems and in-country laboratory capacity to facilitate more rapid responses to future YFV outbreaks.

 

Taye Gari’s PhD thesis: Malaria, anaemia and undernutrition in Ethiopia

Taye Gari. Malaria, anaemia and undernutrition in a drought-affected area of the Rift Valley of Ethiopia: Experiences from a trial to prevent malaria.  PhD. University of Bergen, 2018. Bergen

Summary of PhD thesis

Background: In Ethiopia, malaria, anaemia and undernutrition are common childhood health problems. The country is planning to reduce these conditions to a level where they are not a public health problem. Meanwhile, for the success of this aim, a description of the occurrence and interaction of malaria, anaemia and undernutrition could help contribute to design tailored, efficient and effective control strategies. This study was done in the context of malaria prevention trial, which aimed to measure the effect of combining long-lasting insecticidal nets (LLINs) and indoor residual spraying (IRS) in reducing the malaria incidence compared to LLINs or IRS alone. The study area was affected by a serious drought and famine during the first year of the study.

Objectives: The overall aim of this thesis is to improve our understanding of the effect of malaria prevention on anaemia, and to assess the association between malaria, anaemia and undernutrition among children in a drought-affected area in south-central, Ethiopia.

Methods: A cohort of 5309 residents was followed-up for 16 weeks to measure the variations in malaria incidence among villages (Paper I), and the results were used as a baseline to calculate the sample size needed for the trial. We followed a cohort of children aged 6 to 59 months for one year to describe anaemia and changes in haemoglobin (Hb) concentration (Paper II). A cohort of 4468 children was followed-up for 89 weeks to measure the relationship between malaria and undernutrition (Paper III). Weekly home visits and patient self-referral were also used to identify malaria cases. We conducted Hb concentration (once a year) and anthropometry (twice a year) surveys.

Results: In Paper I, we observed a variation in malaria incidence among villages. On the other hand, the insecticide-treated nets ownership was low (27%), with the distance from the lake or river and younger age being the main risk factors for malaria. The findings of this study were used as a baseline to calculate the sample size for the trial. In Paper II, despite the malaria prevention effort in the community, we observed an unexpected increase in anaemia prevalence over the period of a year, which could be due to the drought and famine that affected the area. A higher incidence of anaemia was observed among children with stunting, malaria infection, young age and in poor families. However, no significant difference in anaemia prevalence was observed among the different trial arms (LLIN+IRS, LLINs alone, IRS alone and routine arm). In Paper III, malaria infection was a risk factor for stunting and wasting, although undernutrition was not a risk for malaria infection. Furthermore, an increase in the prevalence of stunting, but no significant change in a prevalence of wasting was observed over time.

Conclusions: We showed a large variation in malaria incidence among villages. Conducting trials in a drought-prone area may bring an unexpected challenge. We observed an unexpected increase in anaemia prevalence over a year. There was no significant difference in anaemia prevalence among the trial arms. Moreover, a close follow-up of the nutritional status of children with malaria infection may be needed. There could hence be a need to prioritize villages nearer to the main mosquito breeding sites for malaria control.

Trial registration: PACTR 201411000882128 (8 September, 2014)

The Thesis can be downloaded here

Malaria, anaemia and undernutrition

On Friday June 22, 2018 Taye Gari Anaya will defend his PhD thesis at the University of Bergen. 

The title of his thesis is: “Malaria, anaemia and undernutrition in a drought-affected area of the Rift Valley of Ethiopia: Experiences from a trial to prevent malaria”.

Malaria, anaemia and malnutrition are interconnected, and often coexist in prevalent areas such as in sub-Saharan Africa, including Ethiopia. Although individuals in all ages of life are affected by these conditions, children under the age of 5 represent the most vulnerable group. Ethiopia is planning to reduce these conditions to a level where they are not public health problems Meanwhile, for the success of this aim, a description of the occurrence and interaction of malaria, anaemia and undernutrition could help to design tailored, efficient and effective control strategies.  

This thesis measures the effect of malaria prevention on anaemia reduction, and assesses the association between malaria, anaemia and undernutrition among a cohort of children aged 6 to 59 months old followed in a drought-affected rural area in south-central, Ethiopia.

The study showed a large variation in malaria incidence among villages in the same district. Despite community wide malaria prevention effort, an unexpected increase in anaemia prevalence was observed over a year. Malaria infection was a risk factor for undernutrition, although undernutrition was not a risk for malaria infection. There could be a need to prioritise villages nearer to the main mosquito breeding sites for malaria control. Moreover, a close follow-up of the nutritional status of children with malaria infection may be needed. 

Biographical

Taye Gari Ayana is from Hawassa in south Ethiopia. He completed Master degree in Public Health from Addis Ababa University in Ethiopia. From 2013 he has been affiliated with the Centre for International Health as a PhD candidate. Now he is a lecturer at the School of Public Health, at Hawassa University in Ethiopia. His supervisors were Professor Bernt Lindtjørn, Dr Eskindir Loha and Dr Wakgari Deressa. 

Maternal Mortality and Stillbirths More than Halved

The 2017 Annual report from the Centre for International Health at the University of Bergen writes about the implementation reserach in south Ethiopia:

Making standard delivery services available for all is the most important factor for improving maternal mortality. Furthermore, the study suggests that stillbirths can also be reduced if the health services focus on improved obstetric care.

Professor Bernt Lindtjørn led a project in south-western Ethiopia studying how a number of interventions could reduce maternal mortality and stillbirths. As a result of the research efforts, there was 64% reduction in maternal mortality and a 46% reduction in stillbirths.

Lindtjørn explains that there are several reasons why mothers die in childbirth, but he highlights that the main causes are insufficient numbers of institutions and trained personnel, as well as a lack of access to services (i.e. transportation) as being the most important factors.

The project’s positive results were associated with increasing the numbers of trained staff, increasing the num- bers of institutions undertaking Comprehensive Emergency Obstetric Care (CEmOC), increasing the numbers of mothers having Caesarean sections, increasing in the numbers of antenatal controls and referrals, improving access to health service, decreasing in number of home deliveries and decreasing the use of traditional birth attendants.

The goals for the project were straight forward and included making small hospitals available for rural popu- lations, improving referrals (i.e. from health posts to health centres to hospitals when necessary), undertaking additional training for non-clinical physicians, nurses, midwives, and health extension workers, making essential equipment available, and supervising the health care services on offer.

The project was based on the principle of “learning by doing” and involved short courses in Basic Emergen- cy Obstetric Care (BEmOC) and Comprehensive Emergency Obstetric Care (CEmOC), as well as anaesthesiolo- gy, scrub nursing and Continuing Medical Education (CME). It also included Master and PhD training in Public Health: one PhD and 6 master degrees resulted from the project.

The publications from this project can be viewed here.

Reducing stillbirths in Ethiopia

Lindtjorn B, Mitike D, Zidda Z, Yaya Y. Reducing stillbirths in Ethiopia: Results of an intervention programme. PLoS One. 2018;13(5):e0197708. Epub 2018/05/31. doi: 10.1371/journal.pone.0197708. PubMed PMID: 29847607.

Previous studies from South Ethiopia have shown that interventions that focus on intrapartum care substantially reduce maternal mortality and there is a need to operationalize health packages that could reduce stillbirths. The aim of this paper is to evaluate if a programme that aimed to improve maternal health, and mainly focusing on strengthening intrapartum care, also would reduce the number of stillbirths, and to estimate if there are other indicators that explains high stillbirth rates. Our study used a “continuum of care” approach and focussed on providing essential antenatal and obstetric services in communities through health extension workers, at antenatal and health facility services. In this follow up study, which includes the same 38.312 births registered by community health workers, shows that interventions focusing on improved intrapartum care can also reduce stillbirths (by 46%; from 14.5 to 7.8 per 1000 births). Other risk factors for stillbirths are mainly related to complications during delivery and illnesses during pregnancy. We show that focusing on Comprehensive Emergency Obstetric Care and antenatal services reduces stillbirths. However, the study also underlines that illnesses during pregnancy and complications during delivery still represent the main risk factors for stillbirths. This indicates that obstetric care need still to be strengthened, should include the continuum of care from home to the health facility, make care accessible to all, and reduce delays.

 

The risk for death remains high for patients treated for tuberculosis in Ethiopia

Dangisso MH, Woldesemayat EM, Datiko DG, Lindtjorn B. Long-term outcome of smear-positive tuberculosis patients after initiation and completion of treatment: A ten-year retrospective cohort study. PLoS One. 2018;13(3):e0193396. Epub 2018/03/13. doi: 10.1371/journal.pone.0193396. PubMed PMID: 29529036; PubMed Central PMCID: PMCPMC5846790.

Background: The status of tuberculosis (TB) patients since initiation of treatment is unknown in South Ethiopia. The objective of this study was to assess the long-term outcomes of smear-positive TB patients since initiation and completion of treatment, which includes TB recurrence and mortality of TB patients.

Methods: We did a retrospective cohort study on 2,272 smear-positive TB patients who initiated treatment for TB from September 1, 2002-October 10, 2012 in health facilities in Dale district and Yirgalem town administration. We followed them from the date of start of treatment to either the date of interview or date of death.

Results: Recurrence rate of TB was 15.2 per 1000 person-years. Recurrence was higher for re-treatment cases (adjusted hazard ratio (aHR), 2.7; 95% CI, 1.4-5.3). Mortality rate of TB patients was 27.1 per 1,000 person-years. The risk was high for patients above 34 years of age (aHR, 2.1; 95% CI, 1.2-3.9), poor patients (aHR, 1.3; 95% CI, 1.0-1.8), patients with poor treatment outcomes (aHR, 6.7; 95% CI, 5.1-8.9) and for patients treated at least 3 times (aHR 4.8; 95% CI, 2.1-11.1). The excess mortality occurred among patients aged above 34 years was high (41.2/1000 person years).

Conclusions: High TB recurrence and death of TB patients was observed among our study participants. Follow-up of TB patients with the risk factors and managing them could reduce the TB burden.

Successful midway evaluations for 7 PhD students at Hawassa University

On Thursday April 25, seven PhD students admitted to the joint PhD degree programme between the Hawassa University and the University of Bergen, had their mid-way evaluations. Evaluators were senior staff from both Hawassa University and the University of Bergen.

The midway evaluation has the following goals:

  • to find the status regarding the progress and development of the individual PhD project
  • to give the candidate the possibility to present the whole project for a committee

The following students presented their research:

Alemselam Zebdewos: Preventing iron deficiency anaemia: Evaluation of amaranth grain supplementation to 2-5 years old children in southern Ethiopia, a randomized controlled trial

Samrawit Hailu: Childhood illness and health service utilization in Wonago District, South Ethiopia. A community –based cohort study

Sewhareg Belay: Intimate Partner violence during pregnancy: Prevalence, health effect and knowledge about it in Sidama zone, Southern Ethiopia

Hiwot Hailu: Assessment of school health problems in Gedeo Zone, Southern Ethiopia

Bereket Yohannes: Assessing validity of the ‘Household Food Insecurity Access Scale’, and seasonality in food insecurity and undernutrition in rural Southwest Ethiopia

Mehretu Belayneh: Magnitude, seasonality and spatial distribution of under-nutrition among children aged 6-59 months, Boricha, Southern Ethiopia

Moges Tadesse: Maternal and Neonatal illnesses, its economic burden, and health service utilisation in rural Ethiopia: A community-based prospective cohort study

Improved housing reduces malaria

Getawen SK, Ashine T, Massebo F, Woldeyes D, Lindtjorn B. Exploring the impact of house screening intervention on entomological indices and incidence of malaria in Arba Minch town, southwest Ethiopia: A randomized control trial. Acta Trop. 2018;181:84-94. Epub 2018/02/17. doi: 10.1016/j.actatropica.2018.02.009. PubMed PMID: 29452110

House is the major site for malaria infection where most human-vector contact takes place. Hence, improving housing might reduce the risk of malaria infection by limiting house entry of vectors. This study aimed to explore the impact of screening doors and windows with wire meshes on density and entomological inoculation rate (EIR) of malaria vector, and malaria incidence, and assess the acceptability, durability, and cost of the intervention. The susceptibility status of malaria vector was also assessed.

A two-arm randomized trial was done in Arba Minch Town, southwest Ethiopia. 92 houses were randomly included in the trial. The baseline entomological and malaria prevalence data were collected. The mosquito sampling was done twice per household per month by Centers for Diseases Control and Prevention (CDC) light traps for six months. The baseline prevalence of malaria was assessed by testing 396 (83% of the 447 study participants) household members in all the eligible houses. The 92 houses were then randomized into control and intervention groups using mosquito and malaria prevalence baseline data to make the two groups comparable except the intervention. Then, we put wire-mesh on doors and windows of 46 houses. Post-screening mosquito collection was done in each household twice per month for three months. Each household member was visited twice per month for six months to assess malaria episodes. The frequency of damage to different structure of screening was measured twice. In-depth interview was conducted with 24 purposely selected household heads from intervention group. Speciation of Anopheles mosquito was done by morphological key, and the circum-sporozoite proteins (CSPs) analysis was done using enzyme-linked immunosorbent assay. A generalized estimating equation with a negative binomial distribution was used to assess the impact of the intervention on the indoor density of vectors. Clinical malaria case data were analyzed using Poisson regression with generalized linear model.

Screening doors and windows reduced the indoor density of An. arabiensis by 48% (mean ratio of intervention to control = 0.85/1.65; 0.52) (P = .001). Plasmodium falciparum CSP rate was 1.6% (3/190) in the intervention houses, while it was 2.7% (10/372) in the control houses. The protective efficacy of screening intervention from CSP positive An. arabiensis was 41% (mean ratio of intervention to control = 1.6/2.7; 0.59), but was not statistically significant (P = .6). The EIR of An. arabiensis was 1.91 in the intervention group, whereas it was 6.45 in the control group. 477 participants were followed for clinical malaria (50.1% from intervention and 49.9% from the control group). Of 49 RDT positive cases, 45 were confirmed to be positive with microscopy. 80% (n = 36) cases were due to P. falciparum and the rest 20% (n = 9) were due to P. vivax. The incidence of P. falciparum in the intervention group was lower (IRR: 0.39, 95% CI: 0.2-0.80; P = .01) than in the control group. Using incidence of P. falciparum infection, the protective efficacy of intervention was 61% (95% CI: 18-83; P = .007). 97.9% of screened windows and 63.8% of screened doors were intact after eleven months of installation. Malaria mosquito was resistance (mortality rate of 75%) to the insecticide used for bed nets treatment. Almost all participants of intervention arm were willing to continue using screened doors and windows.

Screening doors and windows reduced the indoor exposure to malaria vectors. The intervention is effective, durable and well-accepted. Hence, the existing interventions can be supplemented with house screening intervention for further reduction and ultimately elimination of malaria by reducing insecticide pressure on malaria vectors. However, further research could be considered in broad setting on different housing improvement and in the way how to scale-up for wider community.

 

Intended and non-intended use of bed nets for malaria protection

Doda Z, Solomon T, Loha E, Gari T and Lindtjørn B. A qualitative study of use of long-lasting insecticidal nets (LLINs) for intended and unintended purposes in Adami Tullu, East Shewa Zone, Ethiopia. Malaria Journal 2018;17:69  https://doi.org/10.1186/s12936-018-2209-5

Abstract

Background  Malaria poses a significant public health threat globally, across Africa and in Ethiopia. The use of long-lasting insecticidal nets (LLINs) is currently a proven prevention mechanism. Evidence is building on what happens to LLINs following mass distribution campaigns, with mixed results from different studies, some reporting very low use for intended purposes, others an encouraging level of using for intended purposes. In Ethiopia, between 2005 and 2015, about 64 million LLINs were distributed through periodic mass campaigns with the aims to achieve 100% coverage and 80% utilization. However, studies from rural Ethiopia showed variable LLINs coverage and utilization rate. The MalTrial Project, a collaborative venture between Hawassa University, Ethiopia and NROAID, Norway, has started a trial project in 2014 in Adami Tullu District of central Ethiopia. Quantitative surveys have established evidence on LLINs ownership and utilization, but the behavioural, sociocultural and socioeconomic dynamics of why LLINs’ use for intended purposes is low or why they are employed for other purposes remained elusive. The present qualitative study, building on the quantitative findings and framework, therefore, attempted to fill gaps in these areas using qualitative methods in selected localities of the district.

Methods  The study employed 7 focus groups, 16 individual interviews and observation to undertake data collection in January 2017. The data were analysed using NVivo Version 11 (QSR International) to transcribe, code and identify themes using thematic analysis approach.

Results  The study found out that certain households were more likely to use nets for intended needs in proper ways; a range of factors, notably socio-cultural and poverty, highly influence users’ ideas about the right ways and decisions to use and care for the nets; knowledge gaps and wrong perception exist regarding the purposes and life cycle of the nets; LLINs are employed for repurposed uses once they are considered non-viable, old, or lose their physical integrity; existence of misuse was acknowledged and understood as wrong; and values about gender roles further shape uses, misuses and repurposed use of the nets.

Conclusions  Behavioural, socio-cultural, economic and ecological conditions coupled with deficiencies in perceived bed net design and distribution policies; weak education, communication and social support structures were important in understanding and accounting for why a low level of intended use and a rampant misuse and repurposed use in Adami Tullu community of Ethiopia. A major nexus to address in order to improve intended use of LLINs lies, first and foremost, in economic poverty and socio-cultural factors that underlie much of the misuse and repurposed use of the nets.

Keywords  LLINs Malaria; Intended uses; Misuses; Repurposed uses; Collateral benefits of LLINs