Identifying and calculating the costs of living with trypanosomiasis.

The costs of living with trypanosomiasis among pastoralists or small-scale mixed farmers are multiple and diverse, because of the multiple functions of livestock in these livelihood systems: milk and meat for a household’s own consumption, cash through sale of livestock or livestock products, manure, draught power and social values. These functions will be ranked very differently in different livestock-producing societies. If an agency considering trypanosomiasis control has not already done so, it is imperative that it obtains at least a general ranking of these functions. This can be done in the course of semi-structured interviews, using cards with drawings of e.g. a joint of meat, a bucket (or gourd) of milk, a plough etc., with the local term written on each. Consensus among different livestock-producers is not to be expected, and there may be differences between e.g. villages higher or lower on the hill, or between men and women, but an overall picture should emerge.

It is then important to identify, carefully and in an open-ended way, the different impacts of trypanosomiasis. Some of the impacts identified in pastoral and small-scale mixed-farming systems are as follows:

• The costs of treatment with trypanocides
• Mortality of adult cattle
• Abortion
• Loss of milk
• Loss of draught power and/or the inability to plough at all in certain areas
• Inability to graze in certain areas
• In ability to market livestock, or lower prices obtained for trypanosomiasis-affected animals

When all the different impacts have been identified, there are two complementary ways to investigate them. The first is largely qualitative, based on ranking the different impacts then exploring in a semi-structured way the extent to which they can be quantified for a typical or average livestock-producer. The second is more structured and quantitative, based on a structured, sampled, recall survey. In practice, these can be combined in different ways, for example on including the most important, and/or the most easily quantifiable, impacts in a structured survey.

Ranking impacts is not a substitute for quantification: it can only tell us that one impact is more important than another, not how much more important. But if impacts are intrinsically hard to quantify, or if there is a problem with livestock-producer recall, it may be more effective than a strict quantitative approach. Low ranked impacts need less care in quantification, and high-ranked but unquantifiable costs can be estimated on the principle of being at least as important as the next most ranked category of cost.

Ranking of impacts can also be done with cards, showing e.g. dead cattle, or a gourd of milk crossed out, with a short written description of the impact. Ranking can be done with groups or with individuals, formally sampled, purposively sampled following a formal or informal wealth ranking exercise, or simply chosen as randomly as possible from those attending interviews.

Following ranking, each impact shown to be important for the community can be further explored, with a view to quantifying it, at least to an order of magnitude. Some idea of herd structure, at least of the relative proportions of cows and draught oxen, will be needed for such an exercise. If secondary data does not exist, such estimates can be obtained in a participatory manner by pile sorting. The pros and cons of carrying out a more structured survey can then be considered. Some notes on the major impact categories follow.

Cost of trypanocides
This is likely to be the most easily quantifiable impact. However, livestock-producers may have difficulties recalling amounts spent over one year, particularly if herd-sizes are large. If the price of trypanocides on the local market is known, the number of times a livestock-producer has treated his cattle, or each adult animal, can be multiplied by the price and used as a proxy. When the total costs of trypanosomiasis are calculated/estimated, it should be remembered that in many areas, livestock-producers may spend a significant amount of time searching for treatment and/or drugs. It may not be worth attempting to quantify this time in monetary terms, but it is nonetheless a real cost.

Mortality
In systems where use of trypanocides is common, mortality of cattle may not be a major form of impact. It will also be relatively easy to quantify, although there may be a problem of attributing mortality to trypanosomiasis rather than to other causes. Mortality can be quantified as the proportion of cattle that die, or the number of cattle belonging to the average livestock-producer that die: in either case it will be useful to establish trypanosomiasis-related mortality using a balance sheet of cattle holdings at the present and 12 months before.

Abortions
Abortions are likely to be ranked by livestock producers as a relatively minor impact, and it may be difficult to get accurate recall of their occurrence. A number of questionable assumptions will also be involved in attaching a monetary value to them. Abortions can nevertheless be included in more quantified surveys of the costs of trypanosomiasis.

Loss of milk
Cows contracting trypanosomiasisduring lactation, even if treated promptly with trypanocides, will suffer a serious reduction in milk yields, perhaps for the remainder of the lactation period, and perhaps to the extent of reducing milk surplus for family consumption or sale to zero, and even affecting milk availability for calves.

Studies with pastoralists and agro-pastoralists in Tanzania and Ethiopia suggest that this impact is ranked relatively low, and it may not be necessary to quantify it for purposes of calculating overall costs of trypanosomiasis. Unless milk is customarily sold, it may also be very difficult to quantify. Nevertheless, estimates of the probability of a cow contracting trypanosomiasis during its lactation period (which will be roughly the product of the frequency of it contracting trypanosomiasis at all and the proportion of the year that cows are thought to lactate), daily yields (perhaps using volume measurements of local containers) and a price for milk can be combined to produce estimated costs.

Loss of draught power
This category of impact in fact includes two sub-categories (which may both be felt in any particular system). In situations of low-to-medium challenge where oxen are used for draught power, trypanosomiasis even when promptly treated will mean that an ox cannot plough (or do other work) for anything between one week and one month. This impact can be quantified by getting local estimates for what that period is, and combining it with an estimate for how long in the year an ox is working, the probability of it contracting trypanosomiasis during that time, and a price for the a day’s work by an ox (which may have to be estimated if there is no market in ox hire.

The second sub-category is the situation where livestock-producers would like to use oxen for ploughing, but are prevented from doing so by a situation of high trypanosomiasis challenge. This sort of impact is probably unquantifiable without massive assumptions which would be meaningless in the context of a small community-based tsetse control programme (though not necessarily in the context of an analysis of tsetse control at a regional or national level).

Inability to graze in certain areas
Localised patterns of tsetse infestation may mean that livestock-producers are unable to graze in certain areas that would otherwise be accessible to them. Again, this sort of impact is probably unquantifiable in a way meaningful to a small community-based tsetse control programme.

Inability to market livestock/lower costs
There may be situations where livestock producers would clearly be able to market livestock were it not for the prevalence of trypanosomiasis, or where they do market trypanosomiasis-stressed animals and demonstrably obtain a lower price for them. Such costs can be quantified, though not necessarily with a lot of confidence.

Following such explorations of the main categories of impact of trypanosomiasis, an agency seeking to implement tsetse control can use estimates to compare the overall costs to the costs of tsetse control. For an example of such an estimation procedure click here.

Alternatively, an agency can seek more exact quantification through a structured survey. The latter has both advantages and disadvantages. A more structured and quantitative approach can avoid errors associated with non-random choice of informants, test assumptions used in qualitative work, highlight significant differences between different categories of livestock producer, and may also be more persuasive to donors. On the other hand, a structured survey may face problems of informant recall and unwillingness to quantify, quite apart from the fact that some major impacts will be intrinsically unquantifiable. Those considering structured surveys should also bear in mind the requirements of skilled labour for questionnaire design, pre-testing, supervision, and data cleaning and entry, and in particular for post-survey analysis.

Further notes on questionnaire design, and possible designs of relevant sections are available here.