Bulletin of the World Health Organization

Difficulties in organizing first indoor spray programme against malaria in Angola under the President’s Malaria Initiative

Martinho Somandjinga a, Manuel Lluberas b & William R Jobin c

a. Ministry of Health, Lubango, Angola.
b. Formerly of RTI International, Washington, DC, United States of America (USA).
c. Blue Nile Associates, 25558 Road N6, Cortez, CO, 81321, USA.

Correspondence to William R Jobin (e-mail: blue.nile@earthlink.net).

(Submitted: 28 February 2008 – Revised version received: 16 January 2009 – Accepted: 26 January 2009 – Published online: 24 July 2009.)

Bulletin of the World Health Organization 2009;87:871-874. doi: 10.2471/BLT.08.052514


Control of malaria in epidemic-prone areas of Africa is difficult, partly because of rapid changes in transmission. Where there are short and erratic rainy seasons, control efforts are especially important in rainy years when malaria is deadly, but can be virtually irrelevant in dry years when there is no transmission. This was illustrated in the first attempt to control malaria in Africa by the President’s Malaria Initiative (PMI), a development programme of the Government of the United States of America.


Malaria is an important disease in Angola, where Portuguese is the official language. Malaria is given as the reason for half of all hospital admissions and one-quarter of all hospital deaths. In a population of 10–15 million people, there are 1.5 million malaria cases reported annually.13

In 2005, under the direction of the US Agency for International Development (USAID), PMI began with a malaria control programme in Angola.2 It is not clear why Angola, a country with rich oil resources, was selected. Even though rising oil revenues generated over US$ 1000 per capita, only US$ 1 per capita was designated for malaria in the national budget.4 In the southern provinces where the project started, there was only one malaria officer and one vehicle. The technical staff of the National Malaria Control Program (NMCP) in the capital city of Luanda consisted of four people. They had no vehicles at their headquarters and only a small laboratory and storage room.

Crude data on malaria distribution were provided by the Ministry of Health in reports received from the provinces through normal administrative channels. USAID selected southern Angola as the target area, based on these administrative reports.

Vigorous objections by the authors to the lack of reliable current data on mosquitoes and malaria for the proposed spray area were overcome by what proved to be false reassurances from USAID that staff from the US Centers for Disease Control (CDC) would collect the data in due time.

USAID decided in advance to use spraying of houses with the synthetic pyrethroid lambdacyhalothrin in a wettable powder formulation with 10% active ingredient as the sole control method. It was to be sprayed at 0.03 mg of active ingredient per square metre of interior wall.5

Using spraying as the only method for malaria control can be fast but is unusual. Normally several methods are used in an integrated strategy.612 At least six key components were included in large malaria control efforts by WHO in Nigeria and the Sudan in previous decades.13

In an effort to rush initiation of the PMI, one of the consultants hired by USAID – the technical director – was sent to Angola in August 2005, within a month of PMI being announced. Because USAID was unable to arrange a meeting with the NMCP in Luanda, the technical director travelled directly to the target area in the south. About 400 local men and women were quickly trained to spray the interior walls of houses.


By most measures the spray programme was efficient. The spraying was accomplished at the ideal time in the malaria transmission cycle; just as very light rains began in early December 2005. Spraying continued to the end of March 2006, reaching houses of half a million people in Huila and Kunene Provinces.

After the spray programme was underway, simple observations by the technical director indicated a general absence of adult malaria mosquitoes throughout the province and even an absence of suitable larval habitats. Also, retrospective analysis of rainfall data from a local airport indicated that southern Angola was in the fifth year of a worsening drought, a likely explanation for the lack of mosquitoes.

The problem

Two factors distorted the administrative reports on malaria: irregular omissions of monthly reports from various health units and lack of laboratory confirmation of the malaria cases. Thus other fever-causing diseases were certainly included in the reporting, such as diarrhoeal diseases and seasonal viral infections. This is a problem in many parts of Africa.

The reported malaria data were re-analysed to remove the impact of erratically omitted monthly reports. Further analysis was restricted to those units that had reported data for every month for the entire period of comparison.

The corrected numbers of cases of reported malaria were compared for 2005 and for 2006 in Humpata and Lubango, the two municipalities that had been sprayed, and for Chibia and Kakondo, two similar municipalities in Huila Province that had not been sprayed.

During the normal malaria transmission period in January, February and March of 2006, after the houses in Humpata and Lubango had been sprayed, the number of malaria cases was 30 429, about 89% of the 34 137 reported for the same months during 2005 when no spraying had been conducted (Table 1).

Furthermore, in adjacent and similar unsprayed municipalities of Chibia and Kakondo, the numbers of malaria cases in 2006 was 15 995, also 89% of the 17 975 cases reported in 2005 (Table 1). Thus there was no detected impact of the spraying on reported malaria cases when sprayed and unsprayed municipalities were compared.


Implementation of this large-scale malaria control effort despite the lack of malaria transmission in southern Angola during 2005–2006 was unique and unfortunate. One fault of this project was the rush to spray. Seldom has a large spray programme been so hurried that pretreatment evaluation was omitted.1013 This mistake might be due to lack of experience of the PMI leadership with the complexities of malaria control in Africa. It might also have been wiser to start PMI in a country with a stronger commitment to malaria control.14 Future PMI projects in Angola would also benefit from greater language proficiency by all USAID and CDC personnel.

Unfortunately the global malaria control effort by WHO is currently being evaluated by the same faulty process that caused this mistake in Angola, using administrative reports to evaluate epidemiological progress.1,2,15 Careful monitoring of rainfall and epidemiology greatly improves the cost-effectiveness of malaria control.16


The first PMI programme to control malaria in Africa failed to have an impact in southern Angola because of hurried and inadequate preparation, based primarily on administrative reports of malaria prevalence (Box 1). ■

Box 1. Lessons learned

  • Preparation is important. Don’t rush the project.
  • Determine real need for the programme before investing time and money.
  • Obtain reliable current data. Don’t rely on administrative reports to provide prevalence data.

Competing interests: None of the authors currently work for PMI or USAID.