Background
Given progress in malaria control in recent years, many control programmes in sub-Saharan Africa will soon be required to strengthen systems for surveillance in order to further drive transmission to zero. Yet few practical experiences are available to guide control programmes in designing surveillance system components in low transmission, pre-elimination, and elimination phases.

Methods
A malaria case investigation programme was piloted for 12 weeks in 2012 in Richard Toll district of northern Senegal. Malaria infections (N = 110) were identified through facility-based passive case detection and investigated within three days. Rapid diagnostic tests (RDT) and a brief questionnaire were administered to 5,520 individuals living within the index case compound or within five neighbouring compounds.

Results
In comparison with family and neighbours, index cases were more likely to be male, age 15–49, and to report travel within the past 15 days that entailed an overnight stay. Twenty-three (0.4%) of family/neighbours were RDT-positive. Potential risk factors for infection among family and neighbours were examined, including: sex, age, occupation, travel history, bed net usage, and residence (index vs neighbouring compound). Adjusting for all factors, relative risk (RR) of infection was associated with residence in the index case household (RR = 3.18, p < 0.05) and recent travel, including travel to Dakar (RR = 19.93, p < 0.001), travel within the region (RR = 9.57, p < 0.01), and to other regions in Senegal (RR = 94.30, p < 0.001). Recent fever among RDT-positive family/neighbours was uncommon (30%). Modifications to testing criteria were examined to optimize the efficiency of secondary case investigations in this population. Limiting blood testing to residents of the index case compound and neighbours with recent travel or fever would have identified 20/23 (87%) of the infections through testing 1,173 individuals. Information on the remaining three infections suggests that additional screening for boarding school attendees may facilitate identification of all cases.

Conclusions
The primary risk factor for malaria infection in the low transmission district of Richard Toll is travel. Additional intervention and monitoring strategies to target travellers at risk of malaria infection are needed in this region. Optimizing case investigation with specific targeted testing and treatment of at-risk family and neighbours strengthens the systems needed for continued progress towards malaria elimination in northern Senegal.