The world's scientific and social network for malaria professionals
Subscribe to free Newsletter | 11147 malaria professionals are enjoying the free benefits of MalariaWorld today

Drought and Malaria in Angola

November 26, 2014 - 18:47 -- William Jobin


Drought in southern Angola 2000 to 2006

When I was asked by the US Agency for International Development to go to Angola in 2005 to start the Presidential Malaria Initiative (PMI), I was told to begin spraying interiors of homes in the southern provinces of Huila and Kunene as soon as possible. I think they picked me because Portuguese is one of my favorite languages, and I had worked on malaria control in Sudan for 5 years, besides being with CDC in Puerto Rico when the island was finally declared malaria-free.

However I had never worked in Angola before, and was unable to get good data on malaria or rainfall because of the prolonged civil war which had crippled most government institutions. There was virtually no data in the published scientific literature either, but there was a general feeling among the local staff that the southern provinces were in the zone most prone to malaria epidemics, and the malaria usually occurred around December. There were a lot of contradictory opinions, but no solid data.

When did malaria occur in Angola, and where ?

I recommended to the people in PMI that we should first do a year or two of baseline epidemiology to determine where and when malaria was actually being transmitted, before we tried to stop it. For practical reasons we needed a basis for deciding what months were best to do the spraying. Biocides have a limited residual life of a few months, so usually it is best to spray just before the malaria transmission season, in order to have the longest possible effect on transmission. This is simply a sensible way to get the most impact from the spraying. I was also concerned about getting to remote villages during the rains, when roads were impassable. For that reason we might want to spray early. We needed to know about rainfall and malaria as well.

I also asked them why spraying was the only method we were using. In the Sudan we had always used every available method in an integrated strategy. We called it the 'Kitchen Sink Strategy' because we used everything we could find - except the proverbial Kitchen Sink. This included improved surface drainage in and around communities, elimination of swampy breeding sites, community action through health education, improved irrigation systems and practices, larval source management of all kinds, improved health care with rapid case detection and treatment, and screening of houses. But no - I was told - my job was simply to spray. The outfit from which I learned malaria epidemiology - CDC of Atlanta - was going to do the entomology and epidemiological evaluation. I was told emphatically - spray, and spray quickly.

It seems that the PMI folks in Washington were in a big rush, for political reasons. They wanted to make glowing reports about how quickly they had stopped the malaria in Angola. The PMI was initiated by President George W. Bush in July 2005. So I trained about 400 people and we started spraying in December. The training was really fun, and the spray campaign went like clock work, thanks to Manuel LLuberas and Joaquim Canelas who helped me, as well as Martin Somandjinga and Felix Januario of the Huila Province MOH.

Proudly then, in January, PMI reported to Congress that just a few months after starting the PMI we had already sprayed the homes of half a million people in Angola, showing how efficient PMI was. Unfortunately, that was not the whole story.

Although I was told that somebody from CDC was going to do the entomology, by January they had not shown up, so I did some myself. Happily, I found that there were no adult mosquitoes in the towns we had sprayed. However I was surprised to find that there were no mosquitoes even in the towns we had not sprayed! Nor were there mosquito larvae in the small streams anywhere in the province, nor in the small irrigation reservoirs and canals. It didn’t take long for me to realize that the little surface water I could find was severely polluted, and would not harbor mosquito larvae of any species.

So I tried to do some retrospective analysis on the numbers of malaria cases reported by the local health units during previous years. Actually all I could find were reports on fevers. The health units had no microscopes or diagnostic materials, and malaria and fever were the same word in the local idiom. There were many reporting problems with the data, but after finessing as much as I could, I determined that our spraying program had no impact whatsoever, even on fevers. Finally when I returned to the USA, I was able to get some decent rainfall data for surrounding countries. I learned that the entire region around southern Angola was in the fifth year of a severe drought. The streams and most surface waters had dried up a few years before we arrived.

The lesson, especially for PMI

It was painfully clear that there had been no mosquitoes when we started the spraying, and no malaria. The lesson was obvious and basic; you need pre-treatment data on malaria and rainfall before starting a new program to suppress malaria. And for good measure you should also have comparative data from similar areas where you are not going to spray. These are basic epidemiological evaluation approaches, which CDC seemed to have forgotten.

I was asked to continue with PMI and work in Senegal and Tanzania, the next countries on the list. But I was again told that other people would handle the epidemiology and control operations. My job would be just to spray. Abd whatever epidemiology they were doing it did not involve pre-treatment monitoring of rain and malaria, nor were there going to be treated-untreated comparisons. So I resigned.

But I hope that someday the PMI folks learn the same lesson I did about drought, rain and malaria.


Guy Reeves's picture
Submitted by Guy Reeves on

Dear Bill

While I certainly do not miss the central point of your column, that in some circumstances failing check for mosquitoes, can lead to misdirecting resources.
However, as I understand it , while droughts generally lead to reduce mosquito reproduction, they can in some circumstances increase it.
If I recall correctly several severe malaria epidemics can be linked to droughts e.g. in Sri Lanka 1900-1940s. This drought effect was I think ascribed to the flow in streams ceasing, resulting in the formation of large numbers pools ideal for vector reproduction.

Of course the locations and time window in which this 'drought effect' occurs might be very limited, however should´t it be kept in mind during droughts?


MPI, Plön (Germany)

William Jobin's picture
Submitted by William Jobin on

Thanks Guy,

That is an important point about anopheline species which breed in small pools, but cannot breed in a flowing stream.

So we should add to the urging about collecting rainfall data, to urge malaria control programs to understand the ecology of their vectors. In Malaysia the vectors breed in slow streams as well.


William Jobin Director of Blue Nile Associates