Acknowledgements & background to the report

This report was made possible through the enthusiasm of by Dr Charles Delacollette (WHO, Geneva) and funds from the RBM Resource Network recently created to support the study and management of epidemic malaria in Africa. It was recognised that a thorough review of malaria epidemics from a single country would provide insights into their epidemiological significance and genesis and that such information would provide a more evidence-based platform for the future prediction, prevention and management of epidemics in Africa.

The compilation of such a report has only been made possible through the efforts of many people. In particular the authors would wish to acknowledge Mr Godwin Muranga, a history graduate, who spent many hours at the Kenya National Archives; Mr Joseph Otieno who travelled many miles to transcribe hospital admission data, from often well hidden registers located at district hospitals; Sister Anne Lelesiit (Wamba Mission Hospital) who provided additional information from Samburu district; Mr Sheik Noor Mohammed, Medical Records Officer Wajir, who persevered with RWS in finding the medical registers during the epidemic in Wajir; Dr Beth Rapuoda, Manger National Malaria Control Programme, for her continued support and unique insights into the problems and successes associated with epidemic control during the late 1990's; and Dr Dennis Shanks (Walter Reed, US Army Project, Kenya) for his continued interest in the history of malaria in Kenya and reviewing earlier versions of this report. In addition the compilation of district hospital data was made possible only with the co-operation of District Medical Officers and the Medical Records Officers at each hospital. Finally this report has been developed with a wider context of collaborative research in Nairobi funded by the Wellcome Trust and the Ministry of Health, Government of Kenya both of whom the authors would like to acknowledge.

Contents

1.   Defining epidemics and the scope of the report

1.1 Defining epidemics
1.2 The scope of the present report
1.3 References for Chapter

2.   The distribution of epidemic malaria in Kenya with special reference to clinical patterns

2.1. Overview of Kenya's topography and population
2.2.Climate and altitude determinants of malaria in Kenya
2.3 Epidemic prone districts as defined by Ministry of Health
2.4 Clinical patterns and distribution of disease in selected epidemic districts
2.4.1. Kirinyaga District
2.4.2. Nyamira District
2.4.3. Samburu District
2.4.4. Uasin Gishu District
2.5 References for Chapter

3. The History of malaria epidemics in Kenya: their political significance and control

3.1 1900-1919: The medical services and malaria
3.2 1920 to 1925
3.3 The epidemics of the late twenties
3.4 Malaria as a social disease: a historical review
3.5 The 1930's
3.6 The 1940's
3.7 The eradication era
3.8 Malaria epidemics from the 1990's
3.9 References for Chapter 3

4. The Highlands: the example of Nandi district with reference to neighbouring Kericho district

4.1 Location and population
4.2 The emergence of malaria in Nandi
4.3 Epidemic control in Nandi district
4.4 Changing disease patterns in Nandi district since 1950
4.5 The Kericho example
4.6 The Brookebond tea estate population at Kericho district
4.7 Climate data
4.8 The distribution of infection and vectors in Nandi and Kericho districts
4.9 Summary of the evidence
4.10 References for Chapter 4

5. The arid areas of North Eastern Province: the example of Wajir district

5.1 Location and population
5.2 Climate conditions in NEP
5.3 Health services in Wajir district and the Province
5.4 The Malaria situation in Wajir district
5.5 The 1961-1962 floods
5.6 The El Niño crisis 1997-1998
5.7 The malaria and mortality burden during the epidemic in Wajir district
5.8 Controlling the malaria epidemic in Wajir district
5.9 References for Chapter 5

6. Conclusions: Combining the evidence

6.1 Defining epidemic malaria in Kenya
6.2 Is there a re-emerging epidemic problem Kenya?
6.3. Factors other than climate predisposing new disease patterns in the highlands
6.3.1. Drug resistance
6.3.2 Migration & population mobility
6.3.3 Changing land use patterns
6.4 Control and prevention
6.4.1 Prediction
6.4.2 Organisation of malaria epidemic control
6.4.3 Effective case management
6.4.4 Prophylaxis or MDA
6.4.5 Vector control
6.5 Recommendations for future research
6.6 References for Chapter 6

Appendix A: 1999 Guidelines for malaria epidemic control, MoH, Kenya

Figures

Figure 2.1: Kenya's administrative boundaries
Figure 2.2: Population distribution
Figure 2.3: Climate probability map of stable transmission (Craig et al., 1999)
Figure 2.4: Malaria map of Kenya (Butler et al.,1959)
Figure 2.5: NDVI malaria risk map (Hay et al., 1998)
Figure 2.6: High altitude regions
Figure 2.7: Ministry of Health defined epidemic prone districts
Figure 2.8: Seasonal pattern of malaria admissions of all ages to Kerugoya district hospital, Kirinyaga 1988-1997
Figure 2.9: Malaria admissions to Kerugoya district hospital, Kirinyaga between 1988 and 1997 by age
Figure 2.10: Malaria admissions to Nyamira district hospital from 1990 through to September 1998
Figure 2.11: Malaria admissions to Wamba Mission hospital, Samburu district
Figure 2.12: Ambulatory, microscopically confirmed malaria at Eldoret hospital between 1952 and 1957
Figure 2.13: Malaria admissions to Eldoret district hospital, Uasin Gishu district between 1990 and 1998
Figure 3.1: Outpatient treatments for malaria in Nairobi between 1925 and 1938 (Symes, 1940)
Figure 3.2: European malaria admissions in Nairobi between 1925 and 1938 (Symes, 1940)
Figure 3.3: Notified cases of malaria reported to the Nairobi City Council (De Mello, 1947; Symes, 1940)
Figure 4.1:Location of Nandi and Kericho districts
Figure 4.2: Malaria admissions between 1938 and 1998 at Kapsabet district hospital, Nandi district
Figure 4.3: Monthly childhood admissions to KDH 1986-1998
Figure 4.4: Monthly adult admissions to KDH 1986-1998
Figure 4.5: Ratio of adult-to-child malaria admissions at KDH 1986-1998
Figure 4.6: Malaria admissions to Nandi Hills hospital between 1988 and 1998
Figure 4.7: An. gambiae and An. funestus during 1948 in Buret location, Kericho (Heisch and Harper, 1949)
Figure 4.8: Rate of malaria cases per 1000 population 1942 - 1949, Brookebond tea estates, Kericho (Strangeways-Dixon, 1950)
Figure 4.9: Monthly malaria admissions to Brookebond tea estate hospital, 1965-1998 (Shanks et al., submitted)
Figure 4.10: Rate of malaria admissions to the Brookebond tea estate company, 1986 – 1998
Figure 4.11: Monthly rainfall distribution between 1967 and 1997 in Kericho district
Figure 4.12: Maximum and minimum mean monthly ambient temperatures at Kericho district
Figure 4.13:Location of survey sites for parasitological and entomological data
Figure 4.14: Nandi and Kericho districts: climate suitability model for stable transmission versus malariometric data
Figure 4.15: Nandi and Kericho districts – areas above 1828 metres
Figure 5.1: North Eastern Province – administrative units, infrastructure and population distribution
Figure 5.2: Annual rainfall for the period 1932-1998
Figure 5.3: The monthly rainfall at Wajir between 1991 and 1999
Figure 5.4: Mean monthly maximum and minimum temperatures at Wajir 1991-1999
Figure 5.5: Monthly malaria cases and An.gambiae during the 1943/44 epidemic at Isiolo (Heisch, 1947)
Figure 5.6: Daily Nation, Thursday 12th February 1998
Figure 5.7: Monthly malaria admissions to Wajir district hospital between 1991 and 1999

Defining epidemics

Epidemics invoke fears of exotic and historical diseases, which capture political and global attention. Much of sub-Saharan Africa is exposed to stable, endemic P. falciparum transmission leading to high burdens of morbidity and mortality among young children (Murray & Lopez, 1997; Snow et al., 1999). In addition, the continent has witnessed several devastating malaria epidemics: during the early 1930's in South Africa (Le Sueur et al., 1993), 1958 in Ethiopia (Fontaine et al., 1961), and 1986 in Madagascar (Mouchet, 1998). These outbreaks in transmission followed clearly identifiable changes in climate favouring vector and parasite proliferation and were among non-immune populations. Various estimates from these epidemics indicate that between 1% and 14% of the respective populations died.

Just who is exposed to epidemic conditions and how best to define these conditions are much debated and often hard to quantify. Traditional approaches to measuring whether malaria transmission has the potential for epidemics have focused on the bio-mathematical relationships between the parasite and its primary host, the mosquito. These relationships have been quantified as the Stability Index, Basic Reproduction Rate and Vectorial Capacity.

Christophers' work on the Punjabi epidemics during the early part of this century were some of the first attempts to derive mathematical estimations of ‘epidemic potential' (Anderson & May, 1991). Subsequent models focused on entomological parameters for an index of stability based on the malaria vector's human biting habit and its expectancy of life (MacDonald, 1956):

where "a" is the human biting habit or the average number of human blood meals taken by the female anopheles in a day, 1 / log_ep the life expectancy of the mosquito and "p" is the probability of the vector surviving one day. Values of > 2.5 indicate stability and values of < 0.5 indicate instability.

The Basic Reproduction Rate (R_o ) developed by MacDonald includes duration of sporogony, vector density and duration of infectivity in man. It is therefore a better indicator of stability. The R_o of a parasite is defined as the average number of successful offspring that it is intrinsically capable of producing (Anderson & May, 1991). It is expressed as the average number of secondary infections produced from one infected individual introduced into a non-immune host population. For the parasite to survive successfully R_o must be greater than 1 and transmission becomes unstable when R_o goes below unity. R_o combines measures of mosquito infectivity and survival and is calculated using the following formula:

where ‘p' is the probability of the mosquito surviving through one day, n is the incubation period to infectivity in the mosquito and ‘s,' the proportion of mosquitoes that are sporozoite positive.

The Vectorial Capacity, a mathematical expression of risk of transmission, also includes anopheline density as a risk factor. It is defined mathematically as:

C = ma²pⁿ
      -log_ep

where ‘m' is the relative density of female anophelines, ‘a' the probability that the mosquito will take a human blood meal during a particular day and ‘pⁿ' the proportion of vectors surviving the parasites incubation period (i.e. p - the probability of vector survival and n – the number of days the vector lives). The probability of daily survival is key in determining endemicity levels. For Anopheles gambiae and Anopheles funestus an average daily survival rate of >60% has been shown to be associated with stable endemicity (Wernsdorfer & McGregor, 1988). The Vectorial Capacity is thus a reflection of the mean number of probable inoculations transmitted from one case of malaria in a unit of time and consequently linked to whether transmission in a given area is stable or unstable.

Climatic factors which influence the parameters defined in the formulae presented above for stability and transmission potential have been used in a series of recent models to define the limits of P.falciparum in Africa (Lindsay et al., 1998; Lindsay and Martens, 1998; Craig et al., 1999). These models have provided a series of observations which define geographical areas which constitute low risks of stable endemicity and consequently areas of increased risks of unstable malaria. Epidemic risks, however, may not always coincide with areas defined as unsuitable for stable transmission, or areas of unstable malaria do not always lead to epidemics. These issues are described for Kenya in Chapter 2.

We often classify areas according to whether transmission is stable or unstable. These two extremes are end points in a continuum of differing epidemiological scenarios. The term ‘stable' implies equilibrium. Despite the mathematical simplicity of these concepts the formulae presented above do not provide us with any insights into their effects upon human populations. On the whole where malaria is stable, the prevalence of infection is persistently high and endemicity is relatively insensitive to environmental changes (Molineux, 1988). Variation in transmission is minimal over many years although seasonal fluctuations may occur and transmission can continue even with very few vectors. High levels of immunity develop within the population due to regular and often continuos transmission (MacDonald, 1956). Unstable malaria on the other hand is characterised by great variability in space and time. Collective immunity is low and there is a propensity for epidemics to occur. The disease is also characterised by recession and recurrence and periods when disease incidence is low alternating irregularly with high incidence periods (Warrell & Gilles, 1993).

MacDonald (1957) describes an epidemic as follows

"An epidemic is an acute exacerbation of disease out of proportion to the normal to which the community is subject. There is a proposal to restrict the term to the narrower sense of outbreaks in places where the disease is rare, but the writer [MacDonald] has found this a restricted definition unworkable in practice and prefers the wider and more colloquial term… Epidemics occur only in zones of unstable malaria, where very slight modification in any of the transmission factors may completely upset equilibrium, and where restraining influence of immunity may be negligible or absent, and they therefore show a very marked geographical distribution".

Many factors can influence the ability of parasites and vectors to coexist long enough to result in continued transmission. Several reviews have described the effects on transmission of environmental change, changes in agriculture and forestry practices and man-made construction (Hackett, 1949; Lindsay & Birley, 1996; Lindsay & Martens, 1998; Mouchet et al., 1998). This report will not re-review these particular influences on malaria transmission. However, commentary will be made on the likely significance of small changes in the equilibrium of environmental determinants within the fragile conditions of unstable areas of Kenya.

The main aim of classifying malaria endemicity is to facilitate control and these classifications need to group epidemiological conditions according to common sensitivity to different control options. This review for Kenya will identify some of the weaknesses in our classical approaches to defining stability, epidemics and the need for a review of our classification systems to provide more appropriate public health tools.

1.2. The scope of the present report

Kenya provides an example of how extreme epidemiological patterns of P.falciparum transmission can coexist within a single national boundary. Kenya also has a well documented history of malaria encounters since 1900 across the country co-incidental with the rapid development of colonial, economic expansion and exploitation. We have used a series of archival material held by the Kenya National Archives to understand the clinical and political significance of epidemics between 1900 and 1970. Newspaper articles have provided the popular press's perception of epidemics and how these reports influenced Ministry of Health responses. The old, and often forgotten, literature from local and international journals have been sought to provide quantifiable descriptions of malaria epidemics as they affected different parts and different sectors of Kenya's changing society. And, finally we have visited 7 areas of Kenya affected by the unstable P.falciparum transmission to derive hospital admission data to express the temporal and secular patterns of clinically complicated malaria in each area. Combined, these reports, data and anecdotes have been used to build a picture of how epidemics have affected Kenya's population over this century. The epidemiology, previous attempts at control and their role within a changing health sector have allowed a series of observations and recommendations which should have significance for not only Kenya but those countries in the sub-region with similar ecologies.

1.3: References for Chapter 1.

Anderson RM & May RM (1991). Infectious diseases of humans. Dynamics and control. Oxford University Press.

Craig MH, Snow RW, le Sueur D (1999). African climatic model of malaria transmission based on monthly rainfall and temperature. Parasitology Today, 15: 105-111

Fontaine RE, Najjar AE, Prince JS (1961). The 1958 malaria epidemic in Ethiopia. American Journal of Tropical Medicine & Hygiene, 10: 795-803.

Gilles HM, Warrell DA (1993). Bruce Chwatt's Essential Malariology. Third Edition.

Le Sueur D, Sharp BL, Appleton CC (1993). Historical perspective of the malaria problem in Natal with emphasis on the period 1928-1932. South African Journal of Science, 89: 232-239.

Lindsay SW and Birley MH (1997). Climate change and malaria transmission. Annals of Tropical Medicine & Parasitology, 90: 573-588.

Lindsay SW & Martens WJM (1998). Malaria in the African highlands: past present and future. Bulletin of World Health Organization, 76:33-45.

Lindsay SW, Parson L, Thomas CJ (1998). Mapping the ranges and relative abundance of the two principal African malaria vectors, Anopheles gambiae sensu stricto and An. arabiensis, using climate data. Proceedings of Royal Society. Lond. Series B, 265: 847-854.

MacDonald G (1956). Epidemiological basis of malaria control. Bulletin of the World Health Organisation 15: 613-626

MacDonald G (1957). The epidemiology and control of malaria. Oxford University Press, London.

Molineux (1988) In: Wernsdorfer WH, McGregor I (1988). Malaria: Principles and Practice of Malariology. Volume 2. Churchill Livingstone.

Mouchet J (1998). L'origine des epidemies de paludisme sur les plateaux de Madagascar et les montagnes d'Afrique de l'est et du sud. Bull Soc Pathol Exot 91:64-66

Mouchet J, Manguin S, Sircoulon, Laventure S, Faye O, Onapa AW, Carnevale P, Julvez J, Fontenille D (1998). Evolution of malaria in Africa for the past 40 years: impact of climatic and human factors. Journal of American Mosquito Control Association, 14: 121-130.

Murray CJL and Lopez AD (1997). Mortality by cause for eight regions of the world: Global Burden of Disease Study. Lancet, 349:1269-1276.

Snow RW, Craig MH, Deichmann U, Marsh K. Estimating mortality, morbidity and disability due to malaria among Africa's non-pregnant population. Bulletin of World Health Organisation, in press.

Wernsdorfer WH, McGregor I (1988). Malaria: Principles and Practice of Malariology. Volume 2. Churchill Livingstone.