longer relapse intervals than tropical strains ( Garnham et al., 1975 ; Shute et al.,
1976 ; Cogswell, 1992 ; Collins and Jeffery, 1996 ; Adak et al., 1998 ; Baird et al.,
2007 ; Imwong et al., 2007 ). Primaquine is currently the only widely avail-
able drug with activity on the hypnozoite stage capable of preventing relapse
( Baird and Hoffman, 2004 ; Galappaththy et al., 2007 ), but is associated with
haemolysis in individuals with glucose-6-phosphate dehydrogenase (G6PD)
deficiency ( Beutler, 1994 ; Baird and Hoffman, 2004 ; Cappellini and Fiorelli,
2008 ). Primaquine is contraindicated in pregnant women because of the risk
of haemolytic anaemia in the foetus of unknown G6PD status ( WHO, 2010a ).
The reader is referred to reviews that address relapse (Chapter 2, Volume 80),
treatment (Chapters 4 and 5, Volume 80) and G6PD deficiency (Chapter 4,
Volume 81), which are provided elsewhere in this special issue Advances in
Parasitology . Relapse also has implications for understanding the burden of
P. vivax malaria based on prevalence rates derived from malariometric surveys
and cartographic studies ( Patil et al., 2009 ; Hay et al., 2010b ).
To accurately illustrate P. vivax endemicity, it is necessary to incorporate
the distribution of the Duffy-negative phenotype. The varying prevalence
of Duffy negativity in populations throughout the world is a significant
determinant of the distribution of P. vivax ( Livingstone, 1984 ). Duffy-
negative individuals are, for the most part, refractory to P. vivax infection
and the phenotype is found at highest frequencies in Africa, whereas it is
relatively rare elsewhere ( Howes et al., 2011 ). A detailed review of the effect
of Duffy negativity on the epidemiology of P. vivax (Chapter 2, Volume 81)
is provided elsewhere in this series. The influence of Duffy negativity on
P. vivax transmission reinforces the need to differentiate strategies employed
to generate and interpret maps of P. vivax endemicity from those used for
P. falciparum ( Hay et al., 2009 ; Gething et al., 2011a ).
Until recently, little work had been done to define the geographic lim-
its and risk of P. vivax infection. The only map including vivax malaria
endemicity was that of Lysenko from 1968 ( Lysenko and Semashko, 1968 ).
As shown in Fig. 1.1 , Lysenko defined endemicity as the parasite rate
(PR) in children aged between 2 and 10 years old (hypoendemic <10%,
mesoendemic 11-50%, hyperendemic 51-75%), with the exception of the
holoendemic class (>75%) where the PR was defined in the 1-year age
cohort. The map was derived from a synthesis of historical records and
malariometric indices of all four human malaria parasites: disease and vector
presence and absence records; spleen, parasite, sporozoite and biting rates;
sickle cell incidence; and others ( Hay et al., 2004 ; Gething et al., 2010b ).
Lysenko interpolated the data globally to determine the distribution of