Open Access Open Badges Research

Using the SaTScan method to detect local malaria clusters for guiding malaria control programmes

Marlize Coleman1, Michael Coleman2*, Aaron M Mabuza3, Gerdalize Kok3, Maureen Coetzee45 and David N Durrheim6

Author Affiliations

1 School of Animal, Plant & Environmental Sciences, University of the Witwatersrand, Johannesburg, Gauteng, South Africa

2 Liverpool School of Tropical Medicine, Pembroke Place, Liverpool, L3 5QA, UK

3 Mpumalanga Department of Health, 66 Anderson Street, Nelspruit, 1200, South Africa

4 Vector Control Reference Unit, National Institute for Communicable Diseases, National Health Laboratory Service, 1 Modderfontein Road, Sandringham, 2131 Johannesburg, South Africa

5 SA Research Chair in Medical Entomology & Vector Control, School of Pathology, University of the Witwatersrand, Johannesburg, South Africa

6 Hunter New England Population Health and Hunter Medical Research Institute, Locked Bag 10, Wallsend, 2287, Australia

For all author emails, please log on.

Malaria Journal 2009, 8:68  doi:10.1186/1475-2875-8-68

Published: 17 April 2009



Mpumalanga Province, South Africa is a low malaria transmission area that is subject to malaria epidemics. SaTScan methodology was used by the malaria control programme to detect local malaria clusters to assist disease control planning. The third season for case cluster identification overlapped with the first season of implementing an outbreak identification and response system in the area.


SaTScan™ software using the Kulldorf method of retrospective space-time permutation and the Bernoulli purely spatial model was used to identify malaria clusters using definitively confirmed individual cases in seven towns over three malaria seasons. Following passive case reporting at health facilities during the 2002 to 2005 seasons, active case detection was carried out in the communities, this assisted with determining the probable source of infection. The distribution and statistical significance of the clusters were explored by means of Monte Carlo replication of data sets under the null hypothesis with replications greater than 999 to ensure adequate power for defining clusters.

Results and discussion

SaTScan detected five space-clusters and two space-time clusters during the study period. There was strong concordance between recognized local clustering of cases and outbreak declaration in specific towns. Both Albertsnek and Thambokulu reported malaria outbreaks in the same season as space-time clusters. This synergy may allow mutual validation of the two systems in confirming outbreaks demanding additional resources and cluster identification at local level to better target resources.


Exploring the clustering of cases assisted with the planning of public health activities, including mobilizing health workers and resources. Where appropriate additional indoor residual spraying, focal larviciding and health promotion activities, were all also carried out.