Evidence Action’s Deworm the World Initiative calculates what it costs to deworm a child in the countries where we work. We refer to this as ‘cost per child’ - a number you see in a lot of the literature on the topic. In 2014, we calculated that cost in detail and it’s time to update the numbers, and to transparently and publicly discuss them.  

Why do we calculate the cost of deworming?

In a world of finite resources, both those that implement deworming programs (such as governments) and investors (such as individual and institutional donors) need to have accurate information about the real costs of an intervention to make policy and implementation decisions. We want both policy makers with whom we work, and our investors to have a clear understanding of what it costs in a given country to conduct a large-scale deworming program. We also analyze costs as a management tool to understand where we spend resources and where we can reduce costs.

The costs that we detail below are not an indication of impact, and therefore our cost calculations here are not a cost-benefit analysis - a measure of what it costs for a particular health benefit to be realized.

There are a number of studies that try to assess the cost-effectiveness of deworming such as ‘Worms at Work,” that analyzed the long-run return on investment of mass deworming, and a recent re-analysis of the Cochrane Review, "Does Mass Deworming Affect Child Nutrition?" that takes a close look at weight gain in children because of deworming. It concludes that: "At 0.22 kg per U.S. dollar, the estimated average weight gain per dollar expenditure from deworming MDA (assuming two annual treatments) is more than 35 times that from school feeding programs as estimated in RCTs." 

For the purposes of this cost analysis, however, here we state simply that “for programs supported by Deworm the World, here’s the total cost for each child treated.” 

Because Evidence Action’s Deworm the World Initiative provides technical assistance and is one of a number of partners involved in a deworming program, we also consider costs incurred by other stakeholders such as implementing governments. Similarly, we track the value of donated drugs when these are provided, given that these are an essential input that would otherwise need to be purchased.

What do we include in our cost analysis for deworming and why? 

We follow J-PAL’s 2012 framework that offers guidance on determining costs of large-scale programs in developing countries. We include all partners’ expenditures in determining costs for the the deworming programs, but we do not consider spending that would be incurred even without deworming taking place.

What do we include in our calculation?

  • Teachers’ and principals’ allowances: Our calculation includes per diems and travel allowances paid to teachers and principals when they attend deworming-specific trainings, and when stipulated by local government policies (India no longer pays per diem or travel allowances for teachers or government officials as of 2015). Teachers’ and principals’ general salaries are not included because they do not spend additional time on deworming beyond what they are already compensated for by the government for regular classroom teaching.
  • Drugs: We include the value of drugs in our calculation, whether the drugs are obtained through donation from pharmaceutical companies via the WHO or purchased by the government. Specifically, in India the value of drugs is calculated based on the total amount obtained, as we do not have sufficient confidence that the drugs left over from National Deworming Day are used for treatment through other channels. In Kenya, the cost of drugs is based on the number of children treated multiplied by the value of the drugs used, were the government to purchase them. In Kenya the leftover drugs remain with the Ministry of Health, and are intended for use at local health centers for purposes other than for the National School-Based Deworming Programme.
  • Prevalence surveys: We always include the cost of conducting prevalence surveys in our calculation because they are essential for informing treatment strategies and assessing the long-term impact of deworming. Surveys are done prior to deworming in a given geography to scientifically measure worm prevalence and intensity and then, following multiple rounds of deworming, subsequent surveys are conducted to understand the change in worm prevalence and intensity. Survey frequency varies across programs, but in general, survey costs are amortized across the treatment rounds they inform.
  • Technical implementation assistance: Direct costs such as Evidence Action staff time (both in-country and global) are included in our calculation because these costs are incurred in direct relation to program implementation. Indirect costs are also counted to cover functions like finance and human resources that contribute to our support of the program.
  • Community education and sensitization campaigns: We include all costs of designing and producing promotional materials in local languages. These are used to raise awareness in communities and among parents about deworming, and to encourage treatment participation. They typically include TV or radio ads, village announcers, community parades, and locally-distributed posters and fliers.
  • Training: We include all costs of conducting trainings through a multi-tier cascade to equip government personnel from the national level down to the schools (and sometimes preschools) administering treatment. These include hiring and preparing a team of qualified trainers, developing and printing training materials, and lots of text messages and phone calls for coordination and reminders. When there are specific costs associated with training venues, we include them in the calculation. However, when training (as well as program-related meetings and other events) takes place in spaces that are government-owned and covered by their regular operational costs, those are not included.
  • Monitoring and evaluation: We include all costs of data reporting, monitoring of deworming days, and other critical events, and validating the program coverage and results. This may include: printing of monitoring forms where we manually monitor, costs of developing and using mobile apps where we use mobile phones for data collection, data entry and analysis, travel to schools for monitoring activities, and outsourced or in-house data analysis.

What do we not include? 

  • Maintaining health departments and education systems creates costs that would exist even in the absence of deworming. We do not include these expenses in our analysis but focus instead on the additional costs resulting directly from implementing a deworming program.
  • Following a similar logic, we assume that teachers come to school on deworming day and the same amount of work gets done, in addition to the deworming activity. We thus do not calculate the value of alternative activities that teachers or government officials could be engaging in if not for the deworming program. Although this type of government contribution is not quantified in the model, we note that the existing platforms and infrastructure leveraged by school-based deworming is instrumental to the low cost of these programs.

TL;DR: What is the actual cost of deworming per child for the 2014-2015 school year?

Here are our calculations for the 2014-2015 school year in Kenya, and for the same period for the Indian states of Bihar, Rajasthan and Madhya Pradesh where Evidence Action’s Deworm the World Initiative supported the state governments directly in their deworming programs in that period.

We include in this calculation the total fully-loaded costs of deworming a child in each program location including expenditures from all partners -- Evidence Action, government ministries, the World Health Organization, and others.  We also break down our per-child costs for the program.

Using the total cost per child for each program geography above, the weighted average cost per child treated in the 2014-2015 deworming round was $0.13.  We weight the average to account for the variance in costs and in the number of children treated in each geography.

What has changed in our 2014-2015 cost analysis compared to the previous one? 

In Rajasthan and Bihar, Evidence Action's costs increased by 20% (from $358,191 to $430, 028)  and 16% (from $393,937 to $455,990) respectively between the 2013-2014 and 2014-2015 rounds of deworming.

What accounts for this increase? Even though the share of management costs decreased on a per-child basis as we increased the number of states that we support (another benefit of economies of scale!), we spent more resources on developing high quality public awareness materials (such as radio and TV ads and community banners and posters) than in prior deworming rounds, in line with the Government of India’s National Deworming Day operational guidelines for the 2015 treatment round.

We also incurred higher costs in contracting for independent monitoring services, rather than coordinating this work in-house. Independent monitoring is an important component of our work as it validates key performance indicators such as teacher trainings conducted, drugs delivered, and ultimately children dewormed. These costs reflect what we believe to be a higher quality output.  

During the same time period, the share of costs carried by state governments decreased. In Rajasthan and Bihar, government spending dropped by 44% (from $415,883 to $102,234) and 75% (from $591,886 to $330,864) respectively. This is a result of the introduction of  National Deworming Day guidelines prior to 2014-2015 deworming, that eliminated allowances for teacher trainings as well as for government officials traveling to monitor the program. Reductions in state spending were less dramatic in Rajasthan, where the government increased funding for printing public awareness materials, and hosted deworming day launch events.  

More on the geographic cost variance

The school-based deworming programs we support are implemented by governments and vary greatly in scope. Thus, costs are different from state to state and country to country.  In addition, the costs of doing business vary -- operating costs in Kenya, for example, are much higher than those in India.

Other variations drive costs as well. For instance, the required drug(s) and frequency of their administration vary across countries depending on local worm infection prevalence and intensity. Kenya’s national program treats for two types of parasitic worms: schistosomiasis and soil-transmitted helminths. This requires additional disease mapping and procurement of the more expensive drugs to treat for schistosomiasis. In comparison, India does not have schistosomiasis, and treats only for STH.

Just a reminder here: Our recommendations to governments on treatment frequency follow WHO guidelines that determine how often treatment is administered - typically once or twice a year, depending on scientifically measured worm prevalence.

Costs can be reduced when deworming treatment is administered alongside other health programs. In the Indian state of Bihar, for example, deworming drugs were transported alongside polio vaccines, requiring fewer resources than delivering deworming drugs alone.

Lastly, we are operating within economies of scale. To give you an idea: Bihar, Rajasthan, and Madhya Pradesh are home to more than 30 million at-risk children - more than five times the number of children treated in Kenya.

Why are we not including some geographic areas where we work in this analysis?  

We started working in Ethiopia and in the Indian states of Chhattisgarh, Telangana, and Tripura in late 2015. They are thus not included in the 2014-2015 school year round but will be included in the next cost analysis that will cover the 2015-2016 deworming round. We are not able to provide a complete picture of the costs of the program in Delhi for the 2014-15 round, as we do not have access to the relevant government expenditure data. We have therefore opted to omit Delhi from this analysis.    

How does our cost analysis compare to what others have cited? 

Miguel and Kremer’s 2004 seminal study of a Kenyan deworming program analyzed the costs of school-based deworming, and determined that it was less than $0.50 per child per year.  Alderman’s 2006 study calculates a cost per treatment at $0.21 in Uganda with a cost of $0.42 for two treatments a year, where high prevalence warrants semi-annual deworming.

Since then, Worms at Work (2015), and Owen Ozier's 2015 paper cite a cost of $0.59 per child per year (using much earlier data from the Deworm the World Initiative prior to Evidence Action taking it over, when Kenya was the only program in the portfolio).

As our operating model continues to mature and as we expand into additional geographies, we will continue to partner with governments who contribute to implementation costs at different levels. We also continue to leverage economies of scale and existing school infrastructure, and seek to take advantage of complementary programs.

Our goal is a high level of cost-effectiveness for this proven and effective intervention. The rigorous and regular analysis of program costs allows us to make pragmatic program choices and recommendations to government partners.

No matter where in the world it is implemented, school-based deworming remains a very scalable and extremely low-cost intervention.

 

 

Comment