Insecticide Treated Bed Nets for Malaria Prevention

Malaria is a major cause of preventable mortality in areas endemic to the disease such as Sub-Saharan Africa. Malaria is a mosquito-born disease and prevention of malaria transmission can include vector control measures (such as indoor residual spraying of insecticides and insecticide-treated bed nets) as well as drug-based prevention therapies (such as intermittent prevention therapy during pregnancy). Vulnerable groups to malaria disease include children under five as well as pregnant women (malaria infection during pregnancy can lead to poor health outcomes in infants).

This intervention document focuses on insecicide treated bed nets for malaria prevention.

Abbreviations
Abbreviation	Definition	Note
ITN	Insecticide treated nets
ANC	Antenatal care

Intervention Overview 

Insecticide treated bed nets effectively reduce malaria incidence (and therefore its associated maladies) through the prevention/reduction of mosiquto bites. Children under five and pregnant women are identified as priority groups for the ITN intervention due to their vulnerability to malaria. Due to their high effectiveness and low cost, ITNs have been freely distributed through governmental/other campaigns, as discussed in [Belay-and-Deressa-2008]. However, despite attempts to distribute ITNs, access to the intervention remains limited in some area and a gap between ITN access and use remains [Deressa-et-al-2011]. ITN use during pregnancy has been associated with factors such as household wealth [Deressa-et-al-2014], education [Yitayew-et-al-2018], and ANC attendance [Ouedraogo-et-al-2019].

Todo

Add greater discussion of successes and failures of ITN intervention programs

Affected Outcomes
Outcome	Effect	Modeled?	Note
Malaria Incidence	Decrease	No
Birthweight	Increase in population mean value	Yes	Effect modified by parity in [Gamble-et-al-2007]. Effect entirely mediated through malaria incidence reduction.
Preterm birth	Protective relative risk	No	Statistically insigificant from [Gamble-et-al-2007]. Effect entirely mediated through malaria incidence reduction.
Maternal hemoglobin	Increase in population mean value	No	Statistically insignificant from [Gamble-et-al-2007]. Effect entirely mediated through malaria incidence reduction.

Baseline Coverage Data 

The 2015 Ethiopia Malaria Indicator Survey is a nationally representative survey is a “large, nationally representative survey of coverage of key malaria control interventions, treatment-seeking behavior, and malaria prevalence” (p. 10). The Ethiopia MIS 2015 was cited as the most recent survey data in the 2018 Ethiopia WHO world malaria report. According to the survey, “in malarious areas:” 44% of pregnant women slept under an ITN the previous night and 70% of pregnant women who lived in households that owned at least one ITN slept under an ITN the previous night (Table 21). ITN use among pregnant women was higher in urban areas and among wealthier households. Estiamtes are also available by region. According to the Ethiopia MIS 2015, approximately 60% of the Ehtiopian population lives in Malarious areas.

ITN Baseline Coverage
Location	\(C_\text{malarious areas}\)	\(p_\text{malarious areas}\)	\(C_\text{overall}\)	Note
Ethiopia	0.44 (SD: 0.021), normal distribution of uncertainty. (SD of uncertainty distribution calculated from \(SE = \sqrt{p * (1 - p) / n}\). Data from Ethiopia MIS 2015)	0.6 (data from Ethiopia MIS 2015)	\(C_\text{malarious areas} \times p_\text{malarious areas}\)	Use \(C_\text{overall}\) for simulation coverage proportion

Vivarium Modeling Strategy 

Modeled Outcomes
Outcome	Outcome type	Outcome ID	Affected measure	Effect size measure	Effect size	Note
Birthweight	GBD risk exposure	339	Population mean birthweight	Mean difference	+33 grams (95% CI: 5, 62)	Effect entirely mediated through reduction in malaria incidence

Birthweight 

The ITN intervention affects child birthweight exposures, which are documented here. The intervention should result in an additive change to a simulant’s continuous birthweight exposure value at birth (or upon initialization into the early or late neonatal age groups). We assume there is no corresponding change in a simulant’s gestational age exposure value at birth.

ITN effect on birthweight restrictions
Restriction	Value	Note
Male only	False
Female only	False
Age group start	Birth
Age group end	Late neonatal
Other

ITN and Birthweight Effect Sizes
Population	Effect size	Note
Pregnant women (overall)	+33 grams (95% CI: 5, 62)	[Gamble-et-al-2007]
Pregnant women in first or second pregnancy	+55 (95% CI: 21, 88)	[Gamble-et-al-2007]
Pregnant women in third or later pregnancy	-20 (95% CI: -74, 33)	[Gamble-et-al-2007]

Note

While there is evidence for effect modification of ITN on birthweight by maternal parity, we will model the overall effect until a maternal parity model is developed if/when needed

Todo

Use the distribution of 3rd or later birth order from Ethiopia 2019 DHS

How to sample and apply effect sizes:

Assume a normal distribution of uncertainty within the confidence interval of the effect size in the table above (the code block below describes how to sample from this distribution).
Birthweight exposure values need to be calibrated to baseline ITN coverage in the baseline scenario

from scipy.stats import norm
def sample_from_normal_distribution(mean, lower, upper):
    """Instructions on how to sample from a normal distribution given a mean value and
    95% confidence interval for a parameter"""
    std = (upper - lower) / 2 / 1.96
    dist = norm(mean, std)
    return dist.rvs()

for i in simulants:
  """In the baseline scenario, we need to calibrate baseline coverage
  so that the difference between covered and uncovered babies, on
  average, equals to the effect shift AND that the population mean birthweight value
  from GBD is approximately unchanged.
  * bw_{i} represents the assigned continuous birthweight exposure value for a
  simulant sampled from GBD, which may or may not have already been affected by other
  factors such as maternal BMI, etc. BEFORE consideration of the impact of
  this intervention
  * baseline_itn_coverage represents the baseline coverage proportion"""
  if baseline_itn_coverage_{i} == 'uncovered':
        baseline_supplemented_bw_{i} = bw_{i} - baseline_itn_coverage_{draw} * itn_shift_{draw}
        if alternative_itn_coverage_{i} == 'uncovered':
          alternative_supplemented_bw_{i} = baseline_supplemented_bw_{i}
        elif alternative_itn_coverage_{i} == 'covered':
          alternative_supplemented_bw_{i} = baseline_supplemented_bw_{i} + itn_shift_{draw}
    elif baseline_itn_coverage_ == 'covered':
        baseline_supplemented_bw_{i} = bw_{i} + (1 - baseline_itn_coverage_{draw}) * itn_shift_{draw}
        # makes assumption that all simulants covered in baseline scenario are also covered in alternative scenario
        alternative_supplemented_bw_{i} = baseline_supplemented_bw_{i}

Assumptions and Limitations

We assume that the maternal parity distribution of the study population is similar to that of our modeled population. If the modeled population has a lower parity distribution than the study population, we will underestimate the effect of the distribution (and vise-versa).
Assume that the impact of ITN on birthweight is not mediated through an additional impact in gestational age. As gestational age has an indepedent impact on infant outcomes, this is a conservative assumption.
We are limited in that we do not consider correlation between baseline ITN use and other factors that may be associated with birthweight such as maternal education, maternal age, and ANC attendance.
We assume that malaria burden among the study population in [Gamble-et-al-2007] is similar to the malaria burden among the model population. The [Gamble-et-al-2007] study population included trials performed in Kenya, Ghana, and Thailand in the 1990s. Notably, according to GBD 2019, Ethiopia had substantially lower malaria burden than Ghana, lower burden than Kenya, and substantially greater burden than Thailand at the national level.
We assume that ITNs will impact birthweight among the population living in malarious areas only (60% of the population for Ethiopia). We do not consider differences in birthweight exposure distributions between the populations living in malarious and non-malarious areas.
We assume that there is no effect modification of the ITN intervention by existing use of other malaria control measures such as indoor residual spraying.

Validation and Verification Criteria

In the baseline scenario, the exposure distribution of birthweight (mean birthweight, if available) as well as the mortality rates among the neonatal age groups should match that of GBD.
The coverage of the ITN intervention in the baseline and alternative scenarios should match the associated input values

Child Growth Failure (CGF)

There is little to no evidence that use of ITNs during pregnancy have a direct impact of child growth failure exposure later in life (eg 6-59 months of age), largely due to the logistical challenges of long duration of follow-up. However, there is evidence that birthweight (a directly measured outcome of ITN use during pregnancy) is causally associated with child growth failure exposure later in life. Therefore, we can model an impact of ITN used during pregnancy on child growth failure entirely mediated by its impact on birthweight. We will model this association in the exact same way as described in the child growth failure section of the maternal supplementation intervention document such that the \(S\) shift in birthweight is equal to the total effect of all intervention coverage (or lack of baseline intervention coverage) on birthweight.

Note

ITNs are also recommended for use among children under five in some high burden-malaria settings. While there is evidence that use of ITNs among children under five is associated with decreased malaria burden among this group, there is little to no evidence that there is a direct impact on CGF exposures [Salam-et-al-2014]. Notably, reductions in malaria burden may be associated with reductions in CGF exposure due to a vicious cycle-like pathway between infectious disease and CGF risk; however, we have not included these pathways in our model.

References 

[Gamble-et-al-2007] (1,2,3,4,5,6,7,8)

Gamble, C., Ekwaru, P. J., Garner, P., & ter Kuile, F. O. (2007). Insecticide-treated nets for the prevention of malaria in pregnancy: a systematic review of randomised controlled trials. PLoS medicine, 4(3), e107. https://doi.org/10.1371/journal.pmed.0040107

[Salam-et-al-2014]

Salam, R. A., Das, J. K., Lassi, Z. S., & Bhutta, Z. A. (2014). Impact of community-based interventions for the prevention and control of malaria on intervention coverage and health outcomes for the prevention and control of malaria. Infectious diseases of poverty, 3, 25. https://doi.org/10.1186/2049-9957-3-25

Insecticide Treated Bed Nets for Malaria Prevention

Intervention Overview

Baseline Coverage Data

Vivarium Modeling Strategy

Birthweight