The program for Onchocerciasis Control using community distribution of the drug
Ivermectin has been implemented now since
December '95 in about 19 countries in sub-Saharan Africa.
This program involves state and local health departments,
community groups, NGOs in
training and organizing local villagers to set up their own distribution systems.
And there are a number of levels of data collection that go along with this,
keeping track of village census that the villagers themselves are supposed to do,
keeping records of the numbers of Ivermectin tablets that are given to the distributors,
the distributors keeping record of
how many people receive tablets and how many tablets were dispensed.
And all of this gets reported up.
And so, here we can see that our health information systems provide
us with an opportunity to contemplate whether there is a need for a change.
Some of the things we've seen, for example,
is that there was variation in coverage among different ethnic groups.
Those migrant farm workers who are from a different ethnic group than the local farmers,
the farm laborers, they had less coverage.
They were not participating,
and we wanted to find out why.
There were gender differences.
Now, on this program,
because of the dangers of taking medication during pregnancy,
pregnant women were excluded.
But even when pregnant women were not counted in the denominator,
it turned out that more men than women were receiving the drug at each distribution.
So, by examining these different variations by gender and ethnic group,
one could see that a problem exists and start contemplating,
finding out information of why,
holding focus groups in the community,
talking to women what is happening,
in-depth interviews with some of the late leaders of the farm labor gangs.
These are ways of getting at the problem,
learning more, and being able to go from
the contemplation stage to the preparation or planning stage.
Just because the coverage for this program would go under the stipulated 65 percent,
and it's assumed that if 65 percent take this Ivermectin every year,
for 15 years, Onchocerciasis will be eliminated as public health problem.
There computer simulations should have figured this percentage out.
One of the things discover that there are many reasons
why a low coverage figure might be reported.
One common problem was that the population was being over-counting.
People in the village were counting their relatives who lived in town
and elsewhere in order to make their village look more important.
Again, on the other side,
found that there may be really, in fact,
a low coverage because of poor communication between
the predominantly male leaders of the village
and women's groups and ethnic minorities in the community.
So, again, unless you go further,
you can use your HIS systems to simulate contemplation,
but you still need to go in-depth to find out why.
Ironically, there are even problems of over 85 percent coverage.
Now, the percentage is based on the total population.
If children under five are excluded,
pregnant women at the time of the distribution are
excluded and people who are chronically or severely ill are excluded.
So it's assumed that it would not be possible in an average village
to get much higher than 85 percent if these people were excluded.
And some villages, we got 110 percent coverage.
Another issue, of course, was census problems.
These coverage figures were actually determined by surveys.
The village distributor would submit his or her results in terms of the number of people,
and that would be compared against the village census that they had submitted earlier.
And then as a way of validating this,
teams would go into the village and sample
households to find out coverage through a sample survey.
And what we found, of course,
is that when these reports were 85, 90,
95 percent coverage based on the village distributors reports,
then, of course, go back and say, "Well, what is wrong?"
And it turns out that often it was the census was wrong.
Just as some villages included people who were not resident but were relatives,
in other villages, there were taboos against counting children.
And the total village population should be in
the denominator even if everyone doesn't get the drug.
So, in some places, children were counted,
and it made it look like,
when we did the survey later,
that the number of people treated was much higher than it was.
Then, again, another thing that happened is that people
would stop by from neighboring villages and be counted.
This even occurred in Uganda during times of civil strife in Zaire or Congo.
The people come across the border to get Ivermectin,
and that would inflate the village counts.
All of this goes to show you that it's important to keep health information,
but not just accepted at face value,
but to explore why,
look at the quality of that information and then find out
why people are not accepting the program,
health workers are not performing the way they are.
This again will help your organization help you
contemplate the nature of the organizational problem.
And again, if this contemplation is successful,
it will lead to preparation and planning.
This has implications for the training, for example,
of the community distributors in terms of their understanding of the senses,
their ability to fill out their record forms,
better supervision of the census taking process.
So, again, contemplation is not enough unless it leads to action.
William Brieger, DrPHProfessor, Health Systems Program of the Department of International Health
